Nemko Digital - Members Area

Digital Portfolio

Newsletters

Search

Policy

Author

Area/Country

Policy Type

Details
A Pro-Innovation Approach to AI Regulation

The UK Secretary of State for Science, Innovation and Technology

UK

Policy guidance

Details

Regulations Details

Policy
A Pro-Innovation Approach to AI Regulation

Area/Country
UK

Author
The UK Secretary of State for Science, Innovation and Technology

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://www.gov.uk/government/publications/ai-regulation-a-pro-innovation-approach

Summary
The UK's Department of Science, Innovation and Technology published a policy whitepaper on 29 March 2023, titled ‘A pro-innovation approach to AI regulation’. This document outlines the UK government's plans for a new regulatory framework aimed at fostering innovation in artificial intelligence while ensuring responsible practices. The initiative seeks to boost economic growth and public confidence in AI technologies by addressing potential risks and maintaining fundamental human values. The strategy emphasizes agility, adaptability, and a global leadership role in using AI to address worldwide challenges. It advocates for an evidence-based, continuously adaptive approach to regulation, built around five core principles: safety, transparency, fairness, accountability, and contestability. This framework aims to create an optimal regulatory environment that encourages responsible innovation and mitigates uncertainties.
AB 1651 Workplace Technology Accountability Act

California State Assembly

USA

Regulative proposal

Details

Regulations Details

Policy
AB 1651 Workplace Technology Accountability Act

Area/Country
USA

Author
California State Assembly

Sector
Employment

Policy type
Regulative proposal

Public Private
Private and public sector

Reference link https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202120220AB1651

Summary
Privacy and data protection: The AB 1651 Workplace Technology Accountability Act, proposed in 2022, regulates how employers and vendors collect and use worker data in California. Its primary goal is to safeguard workers' privacy and data rights by introducing regulations on employers' data practices. The bill would grant workers the rights to know, review, correct, and secure their data collected by employers and restrict employers from using this data beyond predefined purposes. It sets out specific limitations on electronic monitoring, the usage of automatic decision systems (ADS), and mandates employers to conduct and disclose impact assessments for technology use. The legislation targets California employers and vendors that collect employee data, monitor electronically, or use automated decision-making tools in worker-related decisions.
AB 331 Automated Decision Tool

California State Assembly

USA

Regulative proposal

Details

Regulations Details

Policy
AB 331 Automated Decision Tool

Area/Country
USA

Author
California State Assembly

Sector
Cross-sector

Policy type
Regulative proposal

Public Private
Private and public sector

Reference link https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202320240AB331

Summary
California Bill AB 331, known as Automated Decision Tools, aims to regulate the application of AI in critical decisions affecting employment and other areas. Starting from 1 January 2025, this proposed legislation requires AI developers to conduct an impact assessment for any automated decision-making tool they create. This assessment must outline the tool's purpose, identify any potential negative impacts, and describe measures to prevent algorithmic discrimination risks. The bill mandates that employees be notified when AI is used in decision-making processes and offers them an opt-out choice. It explicitly prohibits discrimination by algorithms, providing for compensatory damages, declaratory relief, and covering attorney fees and costs for victims of AI-related discrimination. As of now, the bill is under submission and awaits further action.
AI Liability Directive (AILD)

European Commission

EU

Regulative proposal [WITHDRAWN]

Details

Regulations Details

Policy
AI Liability Directive (AILD)

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Regulative proposal [WITHDRAWN]

Public Private
Private and public sector

Reference link https://commission.europa.eu/document/download/f9ac0daf-baa3-4371-a760-810414ce4823_en?filename=1_1_197605_prop_dir_ai_en.pdf

Summary
The AI Liability Directive (AILD), proposed by the European Commission in September 2022, represents a significant step in the European Union's efforts to update its AI liability regulations. This directive is part of a comprehensive AI policy framework that includes the proposed AI Act, aiming to establish fault-based civil liability rules for damages caused by AI systems. The AILD's primary goal is to ensure that individuals affected by AI have the same protection level as those affected by other technologies across the EU. It focuses on harmonizing the rules for non-contractual civil liability and facilitating the process for consumers to claim compensation for damages related to AI products and services. The directive is currently in the legislative pipeline and, once adopted, will require implementation into the national laws of EU Member States.
AI Security Concerns in a Nutshell

German Federal Office for Information Security (BSI)

Germany

Regulative proposal

Details

Regulations Details

Policy
AI Security Concerns in a Nutshell

Area/Country
Germany

Author
German Federal Office for Information Security (BSI)

Sector
Cross-sector

Policy type
Regulative proposal

Public Private
Private and public sector

Reference link https://www.bsi.bund.de/SharedDocs/Downloads/EN/BSI/KI/Practical_Al-Security_Guide_2023.pdf?__blob=publicationFile&v=5

Summary
"AI Security Concerns in a Nutshell," published by the German Federal Office of Information Security in May 2023, is a guideline designed to offer developers an introduction to the primary security challenges associated with AI systems, along with initial defensive strategies. Although not comprehensive, it lays a basic groundwork for understanding AI security. The document categorizes potential attacks into three main types: - Evasion Attacks: These occur during the inference phase of a machine learning model, aiming to cause incorrect classifications. - Information Extraction Attacks: Also known as privacy or reconstruction attacks, these seek to uncover the model or its training data, potentially compromising privacy. - Poisoning and Backdoor Attacks: Poisoning attacks focus on corrupting the machine learning model by tampering with its training data, leading to malfunction or performance degradation. Backdoor attacks, a specific form of poisoning, manipulate the system to produce certain responses when specific conditions are met, without impacting its overall performance during normal operations. The guideline outlines basic defence strategies for mitigating these types of attacks, providing a starting point for developers to secure AI systems against potential threats.
Algorithmic Transparency Recording Standard

Central Digital and Data Office and Centre for Data Ethics and Innovation

UK

Policy guidance

Details

Regulations Details

Policy
Algorithmic Transparency Recording Standard

Area/Country
UK

Author
Central Digital and Data Office and Centre for Data Ethics and Innovation

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Public sector

Reference link https://www.gov.uk/government/publications/algorithmic-transparency-template/algorithmic-transparency-template

Summary
The Algorithmic Transparency Reporting Standard, created by the Central Digital and Data Office (CDDO) and the Centre for Data Ethics and Innovation (CDEI), is a national guideline aimed at increasing transparency within public sector organizations concerning their use of algorithms and the reasons for their deployment. This standard is an integral part of the UK Government's National Data Strategy, focusing on the importance of transparency in decisions supported by algorithms in the public sector. It introduces a two-tiered approach: Tier 1, which provides a basic overview of the algorithmic tool for the public, and Tier 2, which delves into more detailed information about the data and mechanics of the algorithm. The release of the first version of this standard took place in November 2021.
EU Artificial Intelligence Act (AI Act)

European Commission

EU

Regulation

Details

Regulations Details

Policy
EU Artificial Intelligence Act (AI Act)

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Regulation

Public Private
Private and public sector

Reference link https://eur-lex.europa.eu/eli/reg/2024/1689/oj

Summary
The European Union's Artificial Intelligence Act (AI Act) was official published on the 12th of July 2024. The regulation aims to create a comprehensive legal framework applicable to all sectors and types of AI across the EU. The AI Act employs a risk-based approach to regulation, categorizing AI systems into several tiers based on the level of risk they present: - Unacceptable Risk Systems: These are banned within the EU. - High-Risk Systems: Subject to stringent obligations and assessment requirements. - General Purpose and Foundation Models: Tailored requirements are still under discussion, with proposals to ensure they protect fundamental rights and the environment, among other factors. - Limited Risk Systems: Required to meet transparency obligations. - Minimal or No-Risk Systems: Exempt from the regulatory requirements specified in the Act. The legislation primarily focuses on high-risk AI systems, including those involved in critical sectors like biometric identification, critical infrastructure management, law enforcement, and judicial processes. Such systems must comply with several mandatory requirements both pre and post-market, covering risk management, data governance, technical documentation, lifecycle monitoring, accuracy, robustness, cybersecurity, human oversight, user transparency, and quality management. The EU reaches a provisional political agreement on the AI Act on December 8th 2023. Following the finalisation of further details, the AIA is expected to become law in H1 2024. A two year phase implementation period will then begin.
Blueprint for an AI Bill of Rights

The White House

USA

Policy guidance

Details

Regulations Details

Policy
Blueprint for an AI Bill of Rights

Area/Country
USA

Author
The White House

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://www.whitehouse.gov/ostp/ai-bill-of-rights/

Summary
The Blueprint for an AI Bill of Rights, published in October 2022 by the White House Office of Science and Technology Policy (OSTP), is a framework consisting of five guiding principles aimed at shaping the ethical development and deployment of automated systems. Developed in collaboration with academics, human rights groups, the public, and corporations, these principles are designed to ensure that AI technologies align with democratic values, and protect civil rights, civil liberties, and privacy. The principles are as follows: Safety and Effectiveness: AI systems should be reliable and secure. Protection Against Algorithmic Discrimination: AI should be free from unfair bias. Data Privacy: Users' data should be handled with confidentiality and integrity. Notice and Explanation: There should be transparency about AI systems and their decision-making processes. Human Alternatives, Consideration, and Fallback: Users should have the option to opt for human intervention and to challenge AI decisions. Accompanying these principles is a technical companion document that provides detailed guidance for responsibly implementing these standards in the development, design, deployment, and governance of automated systems. This blueprint serves as a reference for stakeholders involved in AI to ensure that technology serves the public good while respecting individual rights and freedoms.
California Consumer Privacy Act (CCPA)

California State Legislature

USA

Regulation

Details

Regulations Details

Policy
California Consumer Privacy Act (CCPA)

Area/Country
USA

Author
California State Legislature

Sector
Cross-sector

Policy type
Regulation

Public Private
Private sector

Reference link https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?division=3.&part=4.&lawCode=CIV&title=1.81.5

Summary
The California Consumer Privacy Act (CCPA) of 2018 is a pivotal law that empowers California residents with greater authority over their personal data held by businesses. It marks a significant step in consumer protection, providing rights such as the ability to inquire about the data collected on them, request deletion of their data, opt-out of their data being sold or shared, and receive protection from discrimination when exercising their CCPA rights. The CCPA, effective from January 2020, was further strengthened by Proposition 24, or the California Privacy Rights Act (CPRA), passed in November 2020, which introduced additional protections like the right to correct inaccurate information and limit the use of sensitive personal data, effective January 1, 2023. The CCPA defines ""business"" as any for-profit entity that collects consumers' personal information, operates in California, and meets specific criteria related to annual revenue, the volume of personal information handled, or revenue generated from selling or sharing personal information. This law encompasses a broad spectrum of businesses, including data brokers. Key responsibilities for businesses under the CCPA include informing consumers about data collection practices, establishing data sharing agreements with third parties, limiting data processing to its initial purpose, and implementing security measures to protect personal data. The CCPA grants Californian consumers the right to be informed about their personal data collection, usage, and sharing; to delete their personal data; to opt-out of the sale or sharing of their personal data; to correct inaccuracies; to restrict the use of sensitive data; and ensures no retaliation for exercising these rights. Businesses are required to comply with these consumer rights, including responding to requests to exercise them and notifying about privacy practices."
Children's Code

Information Commissioners Office (ICO)

UK

Supervisory guidance

Details

Regulations Details

Policy
Children's Code

Area/Country
UK

Author
Information Commissioners Office (ICO)

Sector
Cross-sector

Policy type
Supervisory guidance

Public Private
Private and public sector

Reference link https://ico.org.uk/for-organisations/guide-to-data-protection/ico-codes-of-practice/age-appropriate-design-a-code-of-practice-for-online-services/

Summary
The Children’s Code, also known as the Age-appropriate Design Code, is a data protection guideline established by the UK's Information Commissioners Office (ICO). It aims to ensure that online service providers adhere to data protection laws to safeguard children’s data on the internet. The Code advocates for default settings that inherently protect children’s privacy by limiting data collection and usage, emphasizing the importance of designing online services with the safety and privacy of young users in mind. It is applicable to a broad spectrum of online services that children are likely to access, including apps, online games, and websites and social media platforms. The Code outlines 15 standards that must be met by these services, focusing on practices that prioritize the well-being of child users. Although the Code itself does not carry legal weight, adherence to it is crucial for online services to ensure they comply with the UK's stringent data protection regulations. Published in 2020, the Children’s Code represents a significant step towards enhancing online safety and data protection for children, emphasizing the need for responsible digital environments that cater to the needs and vulnerabilities of younger audiences.
Data Act

European Commission

EU

Regulative proposal

Details

Regulations Details

Policy
Data Act

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Regulative proposal

Public Private
Private sector

Reference link https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM:2022:68:FIN

Summary
The Data Act, proposed in spring 2022 by the European Union, sets out to establish uniform regulations for equitable access to and use of data. This legislation mandates that companies, particularly those manufacturing connected products and providing related services, share the data collected via their products with other parties within the value chain. The primary goal of the Data Act is to promote fair access to both personal and non-personal data, thereby enhancing competition and innovation. Manufacturers and service providers, labeled as ""data holders,"" are required to ensure their products and services are designed in a way that facilitates easy and secure data access for users, including both businesses (B2B) and consumers (B2C). This responsibility extends to sharing data with third parties upon request, broadening the scope of data accessibility across different sectors. The EU institutions finalized the Data Act's text in late June 2023, and it is set to be enforced after a 20-month transition period. This act represents a significant step towards creating a more open and competitive digital economy by ensuring that data, a valuable resource in the modern economy, is accessible to a wider array of stakeholders.
Data Ethics Requirements for RM6200 Artificial Intelligence Suppliers

Crown Commercial Service

UK

Standard

Details

Regulations Details

Policy
Data Ethics Requirements for RM6200 Artificial Intelligence Suppliers

Area/Country
UK

Author
Crown Commercial Service

Sector
Cross-sector

Policy type
Standard

Public Private
Private sector

Reference link https://assets.crowncommercial.gov.uk/wp-content/uploads/Data-ethics-requirements-for-suppliers-template.pdf

Summary
The Data Ethics Requirements for RM6200 Artificial Intelligence Suppliers outline the UK Government's ethical standards for AI suppliers to the public sector. These requirements aim to ensure that the deployment of AI in public decision-making maintains high ethical standards. This initiative is informed by the principles set forth in the Committee on Standards in Public Life's report, ""Artificial Intelligence and Public Standards."" The policy targets suppliers providing AI technologies to UK public sector entities, emphasizing the importance of ethical considerations throughout the adoption of such technologies. Key areas covered include: • System Purpose: Clarifying the objectives of AI systems. • Consent and Control: Managing user consent and control over personal data. • Privacy and Cybersecurity: Ensuring the protection of personal information and system security. • Explainability: Making AI decisions understandable to users. • Ethical Considerations: Addressing ethical issues, including data limitations. • Fairness and Bias: Promoting fairness and mitigating bias within AI systems. • Concept Drift: Handling changes in data over time that might affect AI performance. • Interoperability: Ensuring AI systems can work with other systems and technologies. • Due Diligence: Assessing the ethical implications of existing algorithms or commercial solutions. • Documentation and Auditability: Providing thorough documentation for transparency and facilitating audits. • Lifecycle Management: Managing the entire lifecycle of AI technologies responsibly. • Supplier Skills: Ensuring suppliers have the necessary skills to meet these requirements. This comprehensive framework is designed to guide AI suppliers in the ethical development, deployment, and management of AI services within the UK public sector, ensuring that such technologies are used responsibly and transparently."
Data Governance Act

European Commission

EU

Regulation

Details

Regulations Details

Policy
Data Governance Act

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Regulation

Public Private
Private and public sector

Reference link https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A52020PC0767

Summary
The Data Governance Act is a regulatory framework designed to foster trust in data sharing and increase the availability of data, while also addressing the technical obstacles that hinder the reuse of data. This legislation supports the establishment of common European data spaces in a variety of sectors, including health, environment, energy, agriculture, mobility, finance, manufacturing, public administration, and skills, involving both public and private sector entities. To promote trustworthy data sharing, the Act introduces four main measures: • Enabling the Reuse of Public Sector Data: This provision allows for the wider use of data held by public entities, underlining its potential value beyond its initial collection purpose. • Ensuring Data Intermediaries' Reliability: It establishes the role of data intermediaries as trustworthy organizers of data sharing, ensuring that they facilitate these exchanges in a secure and transparent manner. • Simplifying Data Sharing for Societal Benefit: The Act makes it easier for individuals and businesses to share data, encouraging the use of data for societal good. • Facilitating Cross-Sector and Cross-Border Data Sharing: It aims to remove barriers to data sharing across different sectors and national borders, enhancing the ability to use data effectively and for its intended purpose. The Data Governance Act was enacted on 23 June 2022 and came into full effect on 24 September 2023, marking a significant step towards creating a more integrated and efficient European data economy.
Digital Services Act (DSA)

European Commission

EU

Regulation

Details

Regulations Details

Policy
Digital Services Act (DSA)

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Regulation

Public Private
Private sector

Reference link https://commission.europa.eu/document/download/c33fa400-f978-402f-b530-81e10ff09dd7_en

Summary
The Digital Services Act (DSA) is a landmark regulation by the European Union, designed to create a safer and more transparent online environment while safeguarding consumer and business rights within the EU. It sets forth clear obligations for a broad spectrum of online platforms, including but not limited to internet access providers, hosting services, online marketplaces, app stores, and social networks. Key aims of the DSA include addressing illegal content, combating hate speech and disinformation, and enhancing overall transparency through improved reporting and oversight mechanisms. A significant aspect of the DSA is its focus on fostering innovation, growth, and competitiveness within the EU's digital market. The regulation is notable for its specific attention to very large online platforms and search engines that have a considerable reach within the EU, imposing additional obligations on these entities to ensure a higher level of compliance and to protect users more effectively. The DSA officially came into effect on November 16, 2022, and will be directly applicable across all EU member states, impacting a wide range of online intermediary service providers from January 1, 2024. Some provisions, particularly those related to intermediary services, commenced earlier, signifying the EU's commitment to a more accountable and safer digital space,
Ethics Guidelines for Trustworthy AI

European Commission

EU

Policy guidance

Details

Regulations Details

Policy
Ethics Guidelines for Trustworthy AI

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://ec.europa.eu/newsroom/dae/document.cfm?doc_id=60419

Summary
The Ethics Guidelines for Trustworthy AI, introduced by the High-Level Expert Group on AI on behalf of the European Commission on April 8, 2019, present a comprehensive framework aimed at guiding the development of AI systems to be ethical, secure, and innovative within Europe. The guidelines are structured around the principle that trustworthy AI should meet three critical criteria throughout its lifecycle: legality, adherence to ethical standards, and robustness. This approach ensures that AI systems not only comply with existing laws but also align with ethical values and are resilient against both technical and social challenges, thus minimizing unintentional harm. Targeted at a broad spectrum of AI stakeholders—including but not limited to companies, organizations, researchers, and public services—the guidelines are non-binding and offer best practice advice rather than enforceable rules. They enumerate seven essential requirements for AI to be considered trustworthy: 1. Human agency and oversight 2. Technical robustness and safety 3. Privacy and data governance 4. Transparency 5. Diversity, non-discrimination, and fairness 6. Societal and environmental well-being 7. Accountability These requirements emphasize the importance of AI systems that empower human users, ensure safety and reliability, respect privacy, maintain transparency, promote fairness, benefit society and the environment, and include accountability mechanisms. An assessment checklist is also provided to help stakeholders evaluate AI systems' compliance with these principles. The Ethics Guidelines for Trustworthy AI serve as a practical guide for operationalizing ethical principles in AI development and deployment, reflecting the European Commission's vision for a leading role in ethical AI globally.
European Health Data Space Regulation

European Commission

EU

Regulative proposal [ADOPTED]

Details

Regulations Details

Policy
European Health Data Space Regulation

Area/Country
EU

Author
European Commission

Sector
Health

Policy type
Regulative proposal [ADOPTED]

Public Private
Private and public sector

Reference link https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A52022PC0197

Summary
The European Health Data Space Regulation (EHDSR) proposal, released on May 3, 2022, represents a significant legislative development within the European Union aimed at governing health data. Currently progressing through the legislative procedure, the EHDSR is poised to become a landmark EU regulation upon its enactment. Its primary goals are to bolster patient rights regarding their health data and to increase access to electronic health data repositories. This, in turn, is expected to enhance the data's utilization for the benefit of individuals and the broader population. The EHDSR addresses both primary use of health data for patient care and secondary use for purposes such as research, planning, and policy-making. It builds on the foundational principles laid out in the General Data Protection Regulation (GDPR) and the NIS 2 Directive, integrating these frameworks to regulate the handling of health data securely and ethically.
Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence

The White House

USA

Executive order [REVOKED]

Details

Regulations Details

Policy
Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence

Area/Country
USA

Author
The White House

Sector
Cross-sector

Policy type
Executive order [REVOKED]

Public Private
Private and public sector

Reference link https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/

Summary
Update: This EO, issued on October 30, 2023, was rescinded on January 20, 2025. President Biden issued an Executive Order on 30 October 2023, aimed at navigating the complexities and potential of artificial intelligence. This directive introduces new AI safety and security standards, emphasizes the protection of American privacy, and promotes fairness, civil rights, consumer and worker protection. It also fosters innovation and competition while bolstering American global leadership in AI. This order enhances previous voluntary commitments from leading tech companies towards the development of safe and reliable AI.
General Data Protection Regulation (GDPR)

European Commission

EU

Regulation

Details

Regulations Details

Policy
General Data Protection Regulation (GDPR)

Area/Country
EU

Author
European Commission

Sector
Cross-sector

Policy type
Regulation

Public Private
Private and public sector

Reference link https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32016R0679

Summary
The General Data Protection Regulation (GDPR) is a comprehensive data protection law that came into effect in the European Union (EU) on May 25, 2018. It aims to give individuals control over their personal data while simplifying the regulatory environment for international business by unifying the regulation within the EU. Key provisions of the GDPR include: 1. Consent: Individuals must give clear and explicit consent for their data to be processed, and they have the right to withdraw consent at any time. 2. Right to Access: Individuals have the right to know whether, where, and for what purpose their personal data is being processed, including obtaining copies of their data. 3. Right to Rectification: Individuals can have inaccurate personal data corrected. 4. Right to Erasure: Also known as the ""right to be forgotten,"" this allows individuals to have their data erased under certain conditions. 5. Data Portability: Individuals can request their data in a standard format to use with another service provider. 6. Privacy by Design: Data protection measures must be integrated into product and service development from the outset. 7. Data Protection Officers (DPOs): Organizations processing large amounts of special categories of personal data must appoint a DPO. 8. Breach Notification: Data breaches likely to pose a risk to individuals must be notified to the relevant supervisory authority within 72 hours and to the affected individuals without undue delay. 9. Transfers of Personal Data: Transfers of personal data outside the EU are subject to strict conditions to ensure the level of data protection is not undermined. The GDPR applies not only to organizations within the EU but also to those outside the EU if they offer goods or services to, or monitor the behaviour of, EU data subjects. Non-compliance with the GDPR can cause significant fines up to €20 million or 4% of the company's global annual turnover, whichever is higher. The regulation emphasizes accountability and transparency in data processing, significantly affecting how organizations worldwide handle personal data.
Guidance to Civil Servants on Use of Generative AI

The Cabinet Office

UK

Policy guidance

Details

Regulations Details

Policy
Guidance to Civil Servants on Use of Generative AI

Area/Country
UK

Author
The Cabinet Office

Sector
Government

Policy type
Policy guidance

Public Private
Public sector

Reference link https://www.gov.uk/government/publications/guidance-to-civil-servants-on-use-of-generative-ai/guidance-to-civil-servants-on-use-of-generative-ai

Summary
The "Guidance to Civil Servants on Use of Generative AI," issued on 30 June 2023, sets forth the standards for UK government employees in utilizing Large Language Models in their duties. This directive complements the digital transformation strategy, "Transforming for a Digital Future," with a pledge from central government departments to proactively explore and leverage advancements in technology. Civil servants are urged to foster a curiosity for generative AI, aiming to broaden their comprehension of its applications and operations, while responsibly navigating its potential risks and benefits. The guidance stresses the importance of cautious use of generative AI tools, specifically advising against the processing of sensitive, personal, or confidential government information in compliance with GDPR principles. It highlights the necessity for civil servants to remain alert to the biases and inaccuracies that may arise from AI-generated content, emphasizing the need for thorough verification and proper citation. Additionally, it reiterates the obligation to adhere to GDPR requirements and provides examples of both suitable and unsuitable uses of generative AI within their professional roles. A review of this guidance is scheduled six months post-launch to refine its directives based on new insights and practical experiences with generative AI technology.
Maryland Facial Recognition Law

Maryland General Assembly

USA

Regulation

Details

Regulations Details

Policy
Maryland Facial Recognition Law

Area/Country
USA

Author
Maryland General Assembly

Sector
Employment

Policy type
Regulation

Public Private
Private sector

Reference link https://mgaleg.maryland.gov/mgawebsite/Legislation/Details/HB1202?ys=2020RS

Summary
The Maryland Facial Recognition Law (HB 1202), effective from 1 October 2020, mandates new requirements for employers who utilize facial recognition technology in job interviews. Specifically, it bars the generation of a facial template of an interviewee without obtaining their clear consent through a signed waiver. This document must include the applicant's name, the date of the interview, an acknowledgment of their agreement to the use of facial recognition technology, and a statement confirming their understanding of the waiver.
NIST AI Risk Management Framework (AI RMF)

National Institute of Standards and Technology (NIST)

USA

Policy guidance

Details

Regulations Details

Policy
NIST AI Risk Management Framework (AI RMF)

Area/Country
USA

Author
National Institute of Standards and Technology (NIST)

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.100-1.pdf

Summary
The NIST AI Risk Management Framework (AI RMF) is a set of guidelines designed to help organizations manage risks associated with the deployment and use of artificial intelligence systems. The framework is intended to be adaptable to various sectors and uses of AI, with the flexibility to address a wide range of risk factors, including ethical considerations, fairness, accountability, transparency, and security of AI systems. The AI RMF is part of NIST’s broader efforts to provide foundational principles and a common language for AI risk management. The framework aims to: 1. Provide a structure for organizations to understand and assess the risks associated with AI systems. 2. Help in the design, development, deployment, and operation of AI technologies in a manner that mitigates risks while considering the benefits. 3. Encourage the building of AI systems that are trustworthy, reliable, and governable. 4. Foster an ecosystem that supports responsible AI innovation and use. The framework is typically developed with input from various stakeholders, including industry, academia, and government entities, to ensure a comprehensive approach to AI risk management. It is not a mandatory regulation but serves as a voluntary guide that organizations can choose to follow to ensure their AI systems are developed and used responsibly.
New York City Automated Employment Decision Tool Law

New York City Department of Consumer and Worker Protection

USA

Regulation

Details

Regulations Details

Policy
New York City Automated Employment Decision Tool Law

Area/Country
USA

Author
New York City Department of Consumer and Worker Protection

Sector
Employment

Policy type
Regulation

Public Private
Private sector

Reference link https://rules.cityofnewyork.us/wp-content/uploads/2023/04/DCWP-NOA-for-Use-of-Automated-Employment-Decisionmaking-Tools-2.pdf

Summary
Local Law 144, effective from July 5, 2023, in New York City, introduces regulations governing the use of automated tools in employment-related decisions. The law stipulates that these tools must undergo an annual bias audit and that the results of these audits must be public. Employers and employment agencies in New York City, or those handling remote positions from the city, are mandated to comply with this law when using automated tools to evaluate candidates for hiring or promotion. Automated Employment Decision Tools (AEDTs), as defined by the law, are digital platforms that employ algorithms, including AI, to support or replace human decision-making in employment processes. However, the law does not extend to the use of AEDTs for preliminary processes, such as screening resumes or contacting prospective candidates. An independent entity must execute the required bias audit, which assesses fairness across different demographic groups, and results need to be made accessible to the public. Employers and employment agencies have the responsibility to conduct these audits and cannot delegate it to the tool’s developers. The law mandates that New York City residents be informed of the use of such tools and the criteria they assess with at least 10 days' notice or through a notice on the employer's website and in written policies for current employees.
Medical Device Regulation (MDR)

European Commission

EU

Regulation

Details

Regulations Details

Policy
Medical Device Regulation (MDR)

Area/Country
EU

Author
European Commission

Sector
Health

Policy type
Regulation

Public Private
Private sector

Reference link https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32017R0745

Summary
The Medical Devices Regulation (MDR) is a legislative framework within the European Union that oversees the market introduction, availability, and utilization of medical devices intended for human use, as well as their accessories. This regulation affects manufacturers, distributors, and suppliers, mandating that all medical devices must secure CE marking to demonstrate compliance. Devices are categorized into classes reflecting their purpose and risk levels, with higher-risk devices facing more stringent requirements. The MDR has been in full effect since 26 May 2021.
OECD AI principles

OECD

Worldwide

Policy guidance

Details

Regulations Details

Policy
OECD AI principles

Area/Country
Worldwide

Author
OECD

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://oecd.ai/en/ai-principles

Summary
The OECD AI Principles provide international standards for responsible AI development, emphasizing inclusive growth, human rights, transparency, safety, and accountability. They advocate for AI that contributes to societal well-being and sustainable development, while also promoting fairness and maintaining robust, secure systems. The principles call for accountability from AI system developers and operators. Policy recommendations encourage investment in AI research, fostering a supportive digital ecosystem, preparing the workforce for AI-related changes, and engaging in international collaboration to ensure AI remains trustworthy. These guidelines have been adopted since May 2019 to steer AI use towards positive outcomes.
Recommendation on the Ethics of Artificial Intelligence

UNESCO

Worldwide

Policy guidance

Details

Regulations Details

Policy
Recommendation on the Ethics of Artificial Intelligence

Area/Country
Worldwide

Author
UNESCO

Sector
Education

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449

Summary
UNESCO's Recommendation on the Ethics of Artificial Intelligence is an international standard-setting instrument that aims to ensure AI systems are designed and used in a manner that respects human rights and ethical principles. This recommendation provides a comprehensive framework for the ethical development, deployment, and governance of AI, emphasizing the need for AI to be aligned with the values of respect, protection, and promotion of human dignity. The key aspects of UNESCO's ethical framework for AI include: - Respect for Human Autonomy: AI should support individual autonomy and decision-making, without undermining human agency. - Prevention of Harm: AI systems should be safe, and operate in a way that prevents harm to humans. - Fairness: The development and use of AI should be equitable and inclusive, actively reducing inequality and discrimination. - Transparency and Explainability: AI processes should be transparent, with clear communication about their capabilities and limitations, ensuring individuals understand and can question AI-driven decisions. - Responsibility and Accountability: There should be clear responsibilities and accountability mechanisms for AI systems’ outcomes, including remedial actions in case of adverse impacts. - Privacy and Data Governance: AI must ensure individuals' privacy and data protection, with robust data governance practices that secure personal information. - Knowledge Sharing: The recommendation encourages the sharing of AI knowledge, ensuring that the benefits of AI technologies are widely distributed. - Awareness Raising and Literacy: Efforts should be made to increase public awareness and understanding of AI, fostering digital literacy and informed decision-making. - Societal and Environmental Well-being: AI should be leveraged to promote societal and environmental well-being, contributing positively to sustainable development. - International Collaboration: The recommendation advocates for international cooperation to harmonize AI ethics standards and practices across borders. UNESCO’s recommendation is not legally binding but serves as a guideline for its member states to promote the ethical development and utilization of AI, encouraging them to adopt national regulations and practices that embody these ethical principles.
Responsible Practices for Synthetic Media

Partnership of AI

Worldwide

Policy guidance

Details

Regulations Details

Policy
Responsible Practices for Synthetic Media

Area/Country
Worldwide

Author
Partnership of AI

Sector
Media

Policy type
Policy guidance

Public Private
Private sector

Reference link https://unesdoc.unesco.org/ark:/48223/pf0000381137

Summary
The Partnership on AI (PAI) has established a framework titled "Responsible Practices for Synthetic Media," which sets out guidelines for ethically developing, creating, and distributing synthetic media — content often crafted or altered by AI. This framework targets three key stakeholder groups involved in synthetic media: those building the technology, creators of the content, and entities that distribute or publish it. Central to the framework are the principles of consent, disclosure, and transparency, which guide recommendations on: - Collaboration among stakeholders to deter the harmful use of synthetic media. - Differentiating between responsible and harmful applications of synthetic media. - Implementing targeted mitigation strategies to address the negative impacts when synthetic media is misused. Recognizing the evolving nature of technology and its societal implications, the framework is not static. PAI commits to an annual review of the framework and remains open to additional revisions as suggested by the AI and Media Integrity Steering Committee when necessary.
The Act Concerning Artificial Intelligence, Automated Decision-Making and Personal Data Privacy

Connecticut General Assembly

USA

Regulative proposal

Details

Regulations Details

Policy
The Act Concerning Artificial Intelligence, Automated Decision-Making and Personal Data Privacy

Area/Country
USA

Author
Connecticut General Assembly

Sector
Public services and benefits

Policy type
Regulative proposal

Public Private
Private sector

Reference link https://www.cga.ct.gov/asp/cgabillstatus/cgabillstatus.asp?selBillType=Bill&which_year=2023&bill_num=1103

Summary
The Act Concerning Artificial Intelligence, Automated Decision-Making, and Personal Data Privacy, known as Bill 1103, is a legislative measure adopted by Connecticut to address the use of artificial intelligence (AI) within state agencies. It sets forth a series of deadlines for cataloguing all AI tools used by state entities and mandates the creation of protocols covering the entire lifecycle of AI systems, including their development, procurement, implementation, use, and evaluation. Key components of the bill include the requirement for state agencies to adhere to specific protocols when developing and using automated systems for essential decision-making processes. The bill stipulates the establishment of an artificial intelligence officer position within the Office of Policy and Management, tasked with overseeing the adoption and implementation of procedures related to automated systems by state agencies. The legislation was signed into law by the Governor of Connecticut on June 7, 2023.
The Artificial Intelligence Video Interview Act

Illinois General Assembly

USA

Regulation

Details

Regulations Details

Policy
The Artificial Intelligence Video Interview Act

Area/Country
USA

Author
Illinois General Assembly

Sector
Employment

Policy type
Regulation

Public Private
Private and public sector

Reference link https://www.ilga.gov/legislation/BillStatus.asp?DocNum=2557&GAID=15&DocTypeID=HB&LegID=118664&SessionID=108&SpecSess=&Session=&GA=101

Summary
The Artificial Intelligence Video Act is legislation enacted by the State of Illinois to oversee the use of artificial intelligence (AI) by employers in evaluating video interviews of job applicants. This law applies to any employer leveraging AI tools for analysing video interviews for positions within Illinois. Its main provisions mandate employers must: 1. Notify applicants about the utilization of AI in their assessment process. 2. Clearly explain the criteria and method used by the AI to evaluate their suitability for the job. 3. Candidates must give explicit consent before conducting video interviews that will be analysed using AI. 4. Limit the sharing of video interviews with individuals involved in the hiring process. 5. Delete all video recordings of the applicant within 30 days upon request. This regulation aims to ensure transparency, consent, and privacy in the employment selection process, where AI is used to analysing applicants' video interviews. The government implemented the bill starting January 1, 2020, marking a significant step towards regulating AI applications in hiring practices.
UNICEF Policy Guidance on AI for Children

UNICEF

Worldwide

Policy guidance

Details

Regulations Details

Policy
UNICEF Policy Guidance on AI for Children

Area/Country
Worldwide

Author
UNICEF

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://www.unicef.org/globalinsight/media/2356/file/UNICEF-Global-Insight-policy-guidance-AI-children-2.0-2021.pdf

Summary
The UNICEF Policy Guidance on AI for Children outlines how AI systems can impact children’s rights and well-being. It emphasizes the need for AI to be inclusive, equitable, and to protect children's privacy and safety. The guidance also stresses the importance of transparency, accountability, and empowering both the government and businesses with the knowledge to implement AI responsibly concerning children's rights. It highlights the necessity of preparing children for future developments in AI and create an environment conducive to child-centred AI. The document contains specific recommendations for governments and businesses to follow to ensure that AI systems support and do not undermine children's rights.
NJ A4909

New Jersey State Assembly

USA

Regulative proposal

Details

Regulations Details

Policy
NJ A4909

Area/Country
USA

Author
New Jersey State Assembly

Sector
Employment

Policy type
Regulative proposal

Public Private
Private sector

Reference link https://www.njleg.state.nj.us/bill-search/2022/A4909/bill-text?f=A5000&n=4909_I1

Summary
The legislation outlines rules for entities selling Automated Employment Decision Tools (AEDTs) and for employers implementing them. Sellers cannot offer or advertise AEDTs in New Jersey unless they have conducted a bias evaluation of the tool within the last year or provide a yearly bias audit service to their customers at no extra charge. This audit, which must be performed by an independent party, aims to identify any discriminatory practices against the standards of the New Jersey Law Against Discrimination of 1945. This law outlaws discrimination based on race, creed, color, national origin, or ancestry. Employers utilizing AEDTs must inform applicants, within a 30-day period after the AEDT's application, that the tool was employed during their application process and disclose the criteria evaluated by the AEDT. Unlike New York City's Local Law 144, which mandates prior notification, New Jersey's legislation requires notification post-AEDT use and does not mandate the provision of an alternative evaluation method, even if one is viable. Currently, the bill does not cover employers using AEDTs built in-house. We will need to watch this bill as it develops. The legislation proposes fines up to $500 for the initial violation and up to $1,500 for subsequent violations. These fines are applicable to both the providers of AEDTs who sell these tools in contravention of the regulations (with fines accumulating daily for each AEDT) and to employers who neglect to inform applicants about the application of the AEDT in their evaluation process. The bill has yet to be enacted. According to the latest version, it will become effective immediately upon approval.
NYC A00567

New York State Assembly

USA

Regulative proposal

Details

Regulations Details

Policy
NYC A00567

Area/Country
USA

Author
New York State Assembly

Sector
Employment

Policy type
Regulative proposal

Public Private
Private sector

Reference link https://www.assembly.state.ny.us/leg/?bn=A00567&term=&Summary=Y&Actions=Y&Votes=Y&Memo=Y&Text=Y

Summary
NY A00567 mandates annual bias audits for employers and employment agencies in New York using AI in hiring, focusing on preventing discrimination. It defines Automated Employment Decision Tools (AEDT) broadly, includes a requirement for a disparate impact analysis following the Equal Employment Opportunity Commission guidelines, and emphasizes audits by impartial entities. Unlike similar NYC legislation (NYC LL144), it doesn't require candidate notification or public audit results. There are no specified penalties for non-compliance, but enforcement can involve investigations by the attorney general or commissioner. The bill has not passed yet. Once passes, it would take effect immediately.
Stop Discrimination by Algorithms Act

DC Council

USA

Regulative proposal

Details

Regulations Details

Policy
Stop Discrimination by Algorithms Act

Area/Country
USA

Author
DC Council

Sector
Cross-sector

Policy type
Regulative proposal

Public Private
Private and public sector

Reference link https://lims.dccouncil.gov/Legislation/B24-0558

Summary
The District of Columbia proposed the Stop Discrimination by Algorithms Act, aiming to prevent discrimination by algorithms in vital areas like employment and housing. It prohibits the use of decision-making algorithms based on protected characteristics under the DC Human Rights Act. The Act mandates annual audits, transparency about decision-making processes, and imposes $10,000 fines per violation. It applies immediately upon passage to entities meeting certain criteria, including those with significant data on DC residents or high revenue. Despite initial setbacks, it was reintroduced in February 2023.
Federal AI Governance and Transparency Act

The House Oversight Committee

USA

Regulative proposal

Details

Regulations Details

Policy
Federal AI Governance and Transparency Act

Area/Country
USA

Author
The House Oversight Committee

Sector
Cross-sector

Policy type
Regulative proposal

Public Private
Public sector

Reference link https://oversight.house.gov/wp-content/uploads/2024/03/FedAIGovTransparencyAct.HR7532.Comer_.COMER_029_xml.pdf?utm_source=substack&utm_medium=email

Summary
Chairman James Comer and Ranking Member Jamie Raskin of the House Oversight and Accountability Committee presented the Federal AI Governance and Transparency Act (HR 7532) to the House. This legislation aims to regulate AI's creation, acquisition, use, management, and supervision within the U.S. federal government. It proposes setting federal AI usage standards, enhancing AI policy authority, initiating agency AI governance frameworks, increasing public accountability, eliminating redundant laws, and ensuring alignment with current legislation such as the Privacy Act.
Utah Artificial Intelligence Policy Act (SB 149)

Utah State Legislature

USA

Regulation

Details

Regulations Details

Policy
Utah Artificial Intelligence Policy Act (SB 149)

Area/Country
USA

Author
Utah State Legislature

Sector

Policy type
Regulation

Public Private
Private sector

Reference link https://le.utah.gov/~2024/bills/static/SB0149.html

Summary
The Act assigns legal responsibility and mandates transparency requirements for those deploying generative AI in Utah. It requires that in certain situations, users must be informed when they are interacting with generative AI. Additionally, the law holds generative AI deployers accountable for any legal infractions committed by or with the help of generative AI. Furthermore, the Act creates the Office of Artificial Intelligence Policy, which is responsible for engaging with stakeholders on potential regulatory measures and overseeing the Artificial Intelligence Learning Laboratory Program. The Learning Laboratory Program aims to evaluate the advantages and risks associated with AI technologies within the state, foster the growth of AI in Utah, and assess the impact of, as well as recommend, policies and regulations. The Office is charged with setting a research agenda for the program and is open to applications from individuals or entities interested in participating.
Living Guidelines on the Responsible Use of Generative AI in Research

European Commission

EU

Policy guidance

Details

Regulations Details

Policy
Living Guidelines on the Responsible Use of Generative AI in Research

Area/Country
EU

Author
European Commission

Sector
Research

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://research-and-innovation.ec.europa.eu/document/2b6cf7e5-36ac-41cb-aab5-0d32050143dc_en

Summary
These guidelines help researchers navigate the complexities of AI in academia and broader applications. Key Insight: It's advised to refrain from using generative AI for delicate operations such as peer evaluations and reviews. The integrity of research depends on this cautious approach. The guidelines underscore four foundational principles: - Reliability: Ensuring high-quality research through careful design, methodology, and resource use. - Honesty: Transparent and impartial execution, review, and reporting of research. - Respect: Considering the societal, environmental, and ethical implications of AI use. - Accountability: Holding oneself responsible for the research from conception to publication, including its broader societal impacts. These principles guide the responsible and ethical use of AI tools in research. They are crucial for anyone engaged in AI-related research or fields affected by AI technologies, providing warnings and tools to help ensure responsible innovation.
AI Treaty - The Framework Convention on AI

Council of Europe

EU

Regulation

Details

Regulations Details

Policy
AI Treaty - The Framework Convention on AI

Area/Country
EU

Author
Council of Europe

Sector
Cross-sector

Policy type
Regulation

Public Private
Private and public sector

Reference link https://rm.coe.int/1680afae3c

Summary
The Framework Convention, which is open to countries outside of Europe, establishes a legal framework for the entire lifecycle of AI systems, addressing potential risks while fostering responsible innovation. It employs a risk-based approach to the design, development, usage, and decommissioning of AI systems, considering their potential negative impacts. KEY points The Framework Convention drafts requirements around: ? Fundamental principles: It requires states to ensure AI activities comply with principles like human dignity, equality, privacy, transparency, accountability, and safe innovation. ? Remedies, procedural rights and safeguards: States must document and provide relevant AI system information, allow for decision challenges, enable complaints to authorities, offer procedural guarantees, and notify users of AI interactions. ? Risk and impact management: States must conduct risk and impact assessments, implement mitigation measures, and have the authority to ban certain AI applications. Coverage: The Framework covers both public and private actors. It does not apply to national defence nor R&D activities unless their testing interferes with human rights, democracy, or the rule of law. Monitoring: The Framework will be monitored through the Conference of the Parties to determine the extent to which its provisions are being implemented. They shall also ease co-operation with relevant stakeholders through public hearings.
Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile

National Institute of Standards and Technology (NIST)

USA

Policy guidance

Details

Regulations Details

Policy
Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile

Area/Country
USA

Author
National Institute of Standards and Technology (NIST)

Sector
Cross-sector

Policy type
Policy guidance

Public Private
Private and public sector

Reference link https://airc.nist.gov/docs/NIST.AI.600-1.GenAI-Profile.ipd.pdf

Summary
This document is a companion resource for Generative AI1 2 to the AI Risk Management Framework (AI RMF), pursuant to President Biden’s Executive Order (EO) 14110 on Safe, Secure, and Trustworthy Artificial Intelligence. The document aids in identifying risks that are specific to GenAI, managing them throughout the GenAI lifecycle, and across different sectors. It details 12 risks and 400 actions for mitigation. The document outlines the following risks associated with GenAI: ? Eased access to chemical, biological, radiological, or nuclear (CBRN) weapons ? Confabulation: Producing inaccurate content ? Dangerous Recommendations: Facilitating harmful content and actions ? Data Privacy ? Environmental: Resource-heavy model training impacts. ? Human-AI Configuration: Problems in human-AI interactions. ? Information Integrity ? Information Security ? Intellectual Property ? Obscene Content ? Toxicity, Bias, and Homogenization ? Value Chain and Component Integration
National Artificial Intelligence Policy Framework

Department of Communications and Digital Technologies

South Africa

Policy guidance

Details

Regulations Details

Policy
National Artificial Intelligence Policy Framework

Area/Country
South Africa

Author
Department of Communications and Digital Technologies

Sector

Policy type
Policy guidance

Public Private
Public

Reference link https://techcentral.co.za/wp-content/uploads/2024/08/South-Africa-National-AI-Policy-Framework.pdf

Summary
The National Artificial Intelligence (AI) Policy Framework for South Africa, the initial phase in crafting a comprehensive National AI Policy, aims to integrate AI technologies to boost economic development, improve societal welfare, and establish South Africa as a front-runner in AI innovation. Since the start of the current decade, AI has seen rapid advancements and is now regarded as a general-purpose technology (GPT), akin to electricity or the internet, due to its extensive impact on numerous sectors and its capacity to revolutionize economies and societies. The framework seeks to harness AI’s transformative capabilities to tackle key national challenges and unlock opportunities across all sectors of the economy. Its primary goal is to nurture a dynamic AI ecosystem through a coordinated focus on research and development, talent development, and the strengthening of infrastructure. This comprehensive strategy aims to position AI as a driving force for creating a digital society, fostering a digital economy, and promoting digital inclusion, thereby benefiting all South Africans.
South Korean AI Basic Act

Ministry of Science and ICT

South Korea

Regulation

Details

Regulations Details

Policy
South Korean AI Basic Act

Area/Country
South Korea

Author
Ministry of Science and ICT

Sector
Cross-sector

Policy type
Regulation

Public Private
Public and Private

Reference link https://www.law.go.kr/%EB%B2%95%EB%A0%B9/%EC%9D%B8%EA%B3%B5%EC%A7%80%EB%8A%A5%EB%B0%9C%EC%A0%84%EA%B3%BC%EC%8B%A0%EB%A2%B0%EA%B8%B0%EB%B0%98%EC%A1%B0%EC%84%B1%EB%93%B1%EC%97%90%EA%B4%80%ED%95%9C%EA%B8%B0%EB%B3%B8%EB%B2%95

Summary
The South Korean AI Basic Act sets fundamental principles for AI governance, balancing regulatory oversight with initiatives to enhance innovation and competitiveness. It outlines guidelines for ethical AI use, transparency, and risk management while establishing institutional bodies for oversight and policy development. Main Characteristics of the Law - It defines AI, high-impact AI, generative AI, AI ethics, and AI business operators. - It requires AI business operators to notify users when using high-impact or generative AI. - Operators of systems exceeding a computational threshold must identify, assess, and mitigate risks. - High-impact AI providers must ensure safety and reliability. - The Ministry of Science and ICT can investigate violations and mandate corrective measures. Companies have 1 year to prepare for this regulation

Search

Title

Subject Matter

Type

Status

Audience

AI Aspect

Summary
ISO/IEC 42001:2023

Management system

Published

General, Providers, Developers, User

All

Details

Standards Details

Title
ISO/IEC 42001:2023

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Management system

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Providers, Developers, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/81230.html

Summary
It specifies requirements for establishing, implementing, maintaining, and continually improving an Artificial Intelligence Management System (AIMS) within organizations. This standard is intended for organizations that develop or use AI-driven products or services, ensuring the responsible advancement and use of AI technologies.
ISO/IEC 24027:2021

Bias in AI systems and AI-aided decision making

Published

Developer, Evaluator

Fairness

Details

Standards Details

Title
ISO/IEC 24027:2021

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Bias in AI systems and AI-aided decision making

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
Developer, Evaluator

AI Aspect
Fairness

Buy Standard https://www.iso.org/standard/77607.html

Summary
Addresses bias in AI systems, especially AI-aided decision-making, and describes measurement techniques and methods for assessing bias.
ISO/IEC 24028:2022

Overview of trustworthiness in artificial intelligence

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 24028:2022

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Overview of trustworthiness in artificial intelligence

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/77608.html

Summary
Analyzes factors impacting the trustworthiness of AI systems and surveys existing approaches to support or improve trustworthiness.
ISO/IEC TR 24029-1:2021

Part 1: Overview

Published

Developer, Evaluator

Robustness and Safety

Details

Standards Details

Title
ISO/IEC TR 24029-1:2021

Domain
Artificial Intelligence (AI) – Assessment of the robustness of neural networks

Subject matter
Part 1: Overview

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
Developer, Evaluator

AI Aspect
Robustness and Safety

Buy Standard https://www.iso.org/standard/77609.html

Summary
Provides an overview of approaches available to assess risks related to robustness, focusing on neural networks.
ISO/IEC TR 24030:2024

Use cases

Published

General, Developer, Evaluator, User

N/A

Details

Standards Details

Title
ISO/IEC TR 24030:2024

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Use cases

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
N/A

Buy Standard https://www.iso.org/standard/84144.html

Summary
Illustrates the applicability of AI standardization work across a variety of application domains.
ISO/IEC TR 24372:2021

Overview of computational approaches for AI systems.

Published

General, Developer, Evaluator

N/A

Details

Standards Details

Title
ISO/IEC TR 24372:2021

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Overview of computational approaches for AI systems.

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator

AI Aspect
N/A

Buy Standard https://www.iso.org/standard/78508.html

Summary
Provides an overview of computational approaches for AI systems, describing main algorithms and approaches.
ISO/IEC 38507:2023

Governance implications of the use of artificial intelligence by organizations

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 38507:2023

Domain
Information technology - Governance of IT

Subject matter
Governance implications of the use of artificial intelligence by organizations

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/56641.html

Summary
Provides guidance for the governing body of an organization using or considering the use of AI.
ISO/IEC 22989:2022

Artificial intelligence concepts and terminology

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 22989:2022

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Artificial intelligence concepts and terminology

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/74296.html

Summary
Establishes terminology for AI and describes concepts in the field of AI.
ISO/IEC 23053:2022

Systems Using Machine Learning (ML)

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 23053:2022

Domain
Framework for Artificial Intelligence (AI)

Subject matter
Systems Using Machine Learning (ML)

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/74438.html

Summary
Establishes an AI and ML framework describing system components and their functions in the AI ecosystem.
ISO/IEC 23894:2023

Guidance on risk management

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 23894:2023

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Guidance on risk management

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/77304.html

Summary
Provides guidance on managing risk specifically related to AI, describing processes for effective implementation and integration of AI risk management.
ISO/IEC 5259-1:2024

Part 1: Overview, terminology and examples

Published

Developer

Robustness, Safety, Data Governance

Details

Standards Details

Title
ISO/IEC 5259-1:2024

Domain
Data quality for analytics and machine learning (ML)

Subject matter
Part 1: Overview, terminology and examples

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
Developer

AI Aspect
Robustness, Safety, Data Governance

Buy Standard https://www.iso.org/standard/81088.html

Summary
Provides means for understanding and associating the individual document of the standard series. Foundation for conceptual understanding of data quality for analytics and ML.
ISO/IEC 5259-2

Part 2: Data quality measures

Under publication

Developer

Robustness, Safety, Data Governance

Details

Standards Details

Title
ISO/IEC 5259-2

Domain
Data quality for analytics and machine learning (ML)

Subject matter
Part 2: Data quality measures

Standard type
Standard

Status
Under publication

Publisher
ISO/IEC

Audience
Developer

AI Aspect
Robustness, Safety, Data Governance

Buy Standard https://www.iso.org/standard/81860.html

Summary
provides a data quality model, data quality measures, and guidance on reporting data quality in the context of analytics and machine learning (ML).
ISO/IEC 5259-3:2024

Part 3: Data quality management requirements and guidelines

Published

Developer

Robustness, Safety, Data Governance

Details

Standards Details

Title
ISO/IEC 5259-3:2024

Domain
Data quality for analytics and machine learning (ML)

Subject matter
Part 3: Data quality management requirements and guidelines

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
Developer

AI Aspect
Robustness, Safety, Data Governance

Buy Standard https://www.iso.org/standard/81092.html

Summary
Specifies requirements and provides guidance for establishing, implementing, maintaining, and improving the quality for data used in analytics and ML.
ISO/IEC 5259-4:2024

Part 4: Data quality process framework

Published

Developer

Robustness, Safety, Data Governance

Details

Standards Details

Title
ISO/IEC 5259-4:2024

Domain
Data quality for analytics and machine learning (ML)

Subject matter
Part 4: Data quality process framework

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
Developer

AI Aspect
Robustness, Safety, Data Governance

Buy Standard https://www.iso.org/standard/81093.html

Summary
Provides general common organizational approaches to ensure data quality for training and evaluation in analytics and ML.
ISO/IEC 5259-5

Part 5: Data quality governance

Under publication

Developer

Robustness, Safety, Data Governance

Details

Standards Details

Title
ISO/IEC 5259-5

Domain
Data quality for analytics and machine learning (ML)

Subject matter
Part 5: Data quality governance

Standard type
Standard

Status
Under publication

Publisher
ISO/IEC

Audience
Developer

AI Aspect
Robustness, Safety, Data Governance

Buy Standard https://www.iso.org/standard/84150.html

Summary
Provides a data quality governance framework for analytics and ML.
ISO/IEC 5338:2023

AI system life cycle processes

Published

General, Developer, Evaluator

All

Details

Standards Details

Title
ISO/IEC 5338:2023

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
AI system life cycle processes

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator

AI Aspect
All

Buy Standard https://www.iso.org/standard/81118.html

Summary
Provides processes that support the life cycle stages of an AI system.
ISO/IEC 5339:2024

Guidance for AI applications

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 5339:2024

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Guidance for AI applications

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/81120.html

Summary
The document offers guidance on artificial intelligence (AI) applications, focusing on stakeholder engagement and the AI application life cycle. It seeks to improve multi-stakeholder communication and acceptance by providing a framework that considers the creation, usage, and impact of AI systems.
ISO/IEC 5392:2024

Reference architecture of knowledge engineering

Published

Developer

N/A

Details

Standards Details

Title
ISO/IEC 5392:2024

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Reference architecture of knowledge engineering

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
Developer

AI Aspect
N/A

Buy Standard https://www.iso.org/standard/81228.html

Summary
Defines a reference architecture of knowledge engineering in AI.
ISO/IEC 5469:2024

Functional safety and AI systems

Published

General, Developer, Evaluator, User

Safety

Details

Standards Details

Title
ISO/IEC 5469:2024

Domain
Artificial Intelligence (AI)

Subject matter
Functional safety and AI systems

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
Safety

Buy Standard https://www.iso.org/standard/81283.html

Summary
Describes properties, related risk factors, methods, and processes for the safety of AI systems.
ISO/IEC 6254

Objectives and approaches for explainability of ML models and AI systems

Under development

General, Developer, Evaluator, User

Transparency

Details

Standards Details

Title
ISO/IEC 6254

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Objectives and approaches for explainability of ML models and AI systems

Standard type
Technical specification

Status
Under development

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
Transparency

Buy Standard https://www.iso.org/standard/82148.html

Summary
Describes approaches for achieving explainability objectives of ML models and AI systems.
ISO/IEC 8183:2023

Data life cycle framework

Published

General, Developer, Evaluator

Data Governance

Details

Standards Details

Title
ISO/IEC 8183:2023

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Data life cycle framework

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator

AI Aspect
Data Governance

Buy Standard https://www.iso.org/standard/83002.html

Summary
Defines the stages and identifies associated actions for data processing throughout the artificial intelligence (AI) system life cycle, including acquisition, creation, development, deployment, maintenance and decommissioning.
ISO/IEC 8200:2024

Controllability of automated AI systems

Published

General, Developer, Evaluator, User

Human Agency and Oversight

Details

Standards Details

Title
ISO/IEC 8200:2024

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Controllability of automated AI systems

Standard type
Technical specification

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
Human Agency and Oversight

Buy Standard https://www.iso.org/standard/83012.html

Summary
Defines a framework for the realization and enhancement of automated AI systems’ controllability.
ISO/IEC 12791.2

Treatment of unwanted bias in classification and regression machine learning tasks

Under development

Developer, Evaluator

Fairness

Details

Standards Details

Title
ISO/IEC 12791.2

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Treatment of unwanted bias in classification and regression machine learning tasks

Standard type
Technical specification

Status
Under development

Publisher
ISO/IEC

Audience
Developer, Evaluator

AI Aspect
Fairness

Buy Standard https://www.iso.org/standard/84110.html

Summary
provides mitigation techniques that can be applied throughout the AI system life cycle in order to treat unwanted bias.
ISO/IEC 12792

Transparency taxonomy of AI systems

Under development

General, Developer, Evaluator, User

Transparency

Details

Standards Details

Title
ISO/IEC 12792

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Transparency taxonomy of AI systems

Standard type
Standard

Status
Under development

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
Transparency

Buy Standard https://www.iso.org/standard/84111.html

Summary
Defines a taxonomy of information elements to assist AI stakeholders with identifying and addressing the needs for transparency of AI systems.
ISO/IEC 24368:2022

Overview of ethical and societal concerns

Published

General, Developer, Evaluator, User

All

Details

Standards Details

Title
ISO/IEC 24368:2022

Domain
Artificial Intelligence (AI)

Subject matter
Overview of ethical and societal concerns

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator, User

AI Aspect
All

Buy Standard https://www.iso.org/standard/78507.html

Summary
This document offers a comprehensive overview of ethical and societal concerns related to AI, including information on principles, processes, and methods in this field. It does not promote any particular set of values or value systems. It also includes an overview of International Standards addressing issues stemming from AI ethical and societal concerns.
ISO/IEC 24668:2022

Process management framework for big data analytics

Published

General, Developer, Evaluator

Robustness, Safety, Data Governance

Details

Standards Details

Title
ISO/IEC 24668:2022

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Process management framework for big data analytics

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator

AI Aspect
Robustness, Safety, Data Governance

Buy Standard https://www.iso.org/standard/78368.html

Summary
The document outlines a framework for creating processes to effectively utilize big data analytics across various industries and sectors. Specifically, it details process management for big data analytics, considering various interconnected process categories, including organization stakeholder processes, competency development processes, data management processes, analytics development processes, and technology integration processes. It describes the processes needed to acquire, describe, store, and manage data at the organizational level to provide big data analytics services.
ISO/IEC 4213:2022

Assessment of machine learning classification performance

Published

Developer, Evaluator

Robustness, Safety

Details

Standards Details

Title
ISO/IEC 4213:2022

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Assessment of machine learning classification performance

Standard type
Technical specification

Status
Published

Publisher
ISO/IEC

Audience
Developer, Evaluator

AI Aspect
Robustness, Safety

Buy Standard https://www.iso.org/standard/79799.html

Summary
The document specifies methodologies for measuring classification performance of machine learning models, systems and algorithms.
ISO/IEC 27563:2023

Best practices

Published

General, Developer, Evaluator

Data Governance, privacy, security

Details

Standards Details

Title
ISO/IEC 27563:2023

Domain
Security and privacy in artificial intelligence use case

Subject matter
Best practices

Standard type
Technical report

Status
Published

Publisher
ISO/IEC

Audience
General, Developer, Evaluator

AI Aspect
Data Governance, privacy, security

Buy Standard https://www.iso.org/standard/80396.html

Summary
The document covers: - A comprehensive assessment of security and privacy for the AI system of interest; - Security and privacy concerns; - Security and privacy risks; - Security and privacy controls; - Security and privacy assurance; - Security and privacy plans. Security and privacy are addressed separately because their analyses can differ.
ISO/IEC 17847

Verification and validation analysis of AI systems

Under development

Developer, Evaluator

Robustness and Safety

Details

Standards Details

Title
ISO/IEC 17847

Domain
Information technology - Artificial Intelligence (AI)

Subject matter
Verification and validation analysis of AI systems

Standard type
Technical specification

Status
Under development

Publisher
ISO/IEC

Audience
Developer, Evaluator

AI Aspect
Robustness and Safety

Buy Standard https://www.iso.org/standard/85072.html

Summary
Describes approaches and provides guidance on processes for the verification and validation analysis of AI systems (comprising AI system components and the interaction of non-AI components with the AI system components) including formal methods, simulation and evaluation.
Data Ethics Requirements for RM6200 Artificial Intelligence Suppliers

N/A

Published

Suppliers

Data Governance

Details

Standards Details

Title
Data Ethics Requirements for RM6200 Artificial Intelligence Suppliers

Domain
N/A

Subject matter
N/A

Standard type
Standard

Status
Published

Publisher
Crown Commercial Service

Audience
Suppliers

AI Aspect
Data Governance

Buy Standard https://assets.crowncommercial.gov.uk/wp-content/uploads/Data-ethics-requirements-for-suppliers-template.pdf

Summary
This document outline the UK Government's ethical standards for AI suppliers to the public sector. These requirements aim to ensure that the deployment of AI in public decision-making maintains high ethical standards. This initiative is informed by the principles set forth in the Committee on Standards in Public Life's report, ""Artificial Intelligence and Public Standards."" The policy targets suppliers providing AI technologies to UK public sector entities, emphasizing the importance of ethical considerations throughout the adoption of such technologies. Key areas covered include: • System Purpose: Clarifying the objectives of AI systems. • Consent and Control: Managing user consent and control over personal data. • Privacy and Cybersecurity: Ensuring the protection of personal information and system security. • Explainability: Making AI decisions understandable to users. • Ethical Considerations: Addressing ethical issues, including data limitations. • Fairness and Bias: Promoting fairness and mitigating bias within AI systems. • Concept Drift: Handling changes in data over time that might affect AI performance. • Interoperability: Ensuring AI systems can work with other systems and technologies. • Due Diligence: Assessing the ethical implications of existing algorithms or commercial solutions. • Documentation and Auditability: Providing thorough documentation for transparency and facilitating audits. • Lifecycle Management: Managing the entire lifecycle of AI technologies responsibly. • Supplier Skills: Ensuring suppliers have the necessary skills to meet these requirements. This comprehensive framework is designed to guide AI suppliers in the ethical development, deployment, and management of AI services within the UK public sector, ensuring that such technologies are used responsibly and transparently."
Algorithmic Transparency Recording Standard

N/A

Published

General

Transparency

Details

Standards Details

Title
Algorithmic Transparency Recording Standard

Domain
N/A

Subject matter
N/A

Standard type
Standard

Status
Published

Publisher
Central Digital and Data Office and Centre for Data Ethics and Innovation

Audience
General

AI Aspect
Transparency

Buy Standard https://www.gov.uk/government/collections/algorithmic-transparency-recording-standard-hub

Summary
This document is a national guideline aimed at increasing transparency within public sector organizations concerning their use of algorithms and the reasons for their deployment. This standard is an integral part of the UK Government's National Data Strategy, focusing on the importance of transparency in decisions supported by algorithms in the public sector. It introduces a two-tiered approach: Tier 1, which provides a basic overview of the algorithmic tool for the public, and Tier 2, which delves into more detailed information about the data and mechanics of the algorithm. The release of the first version of this standard took place in November 2021.
IEEE 2801:2022

Recommended Practice for the Quality Management of Datasets for Medical Artificial Intelligence

Published

Developer

Privacy and Data Governance

Details

Standards Details

Title
IEEE 2801:2022

Domain
N/A

Subject matter
Recommended Practice for the Quality Management of Datasets for Medical Artificial Intelligence

Standard type
Standard

Status
Published

Publisher
IEEE

Audience
Developer

AI Aspect
Privacy and Data Governance

Buy Standard https://standards.ieee.org/ieee/2801/7459/

Summary
Highlights quality objectives for datasets in medical AI, describing control of records during the dataset lifecycle.
IEEE 2802:2022

Standard for Performance and Safety Evaluation of Artificial Intelligence Based Medical Devices: Terminology

Published

Developer

Robustness and Safety

Details

Standards Details

Title
IEEE 2802:2022

Domain
N/A

Subject matter
Standard for Performance and Safety Evaluation of Artificial Intelligence Based Medical Devices: Terminology

Standard type
Standard

Status
Published

Publisher
IEEE

Audience
Developer

AI Aspect
Robustness and Safety

Buy Standard https://standards.ieee.org/ieee/2802/7460/

Summary
Establishes concepts for performance and safety evaluation of AI-based medical devices.
IEEE 7000:2021

Standard Model Process for Addressing Ethical Concerns during System Design

Published

Developer

Fairness, Societal and Environmental Wellbeing

Details

Standards Details

Title
IEEE 7000:2021

Domain
N/A

Subject matter
Standard Model Process for Addressing Ethical Concerns during System Design

Standard type
Standard

Status
Published

Publisher
IEEE

Audience
Developer

AI Aspect
Fairness, Societal and Environmental Wellbeing

Buy Standard https://ieeexplore.ieee.org/document/9536679

Summary
Establishes processes to include consideration of ethical values during system design stages.
IEEE 7001:2021

Standard for Transparency of Autonomous Systems

Published

Developer, Evaluator

Transparency

Details

Standards Details

Title
IEEE 7001:2021

Domain
N/A

Subject matter
Standard for Transparency of Autonomous Systems

Standard type
Standard

Status
Published

Publisher
IEEE

Audience
Developer, Evaluator

AI Aspect
Transparency

Buy Standard https://ieeexplore.ieee.org/document/9726144

Summary
Provides a framework to help developers make autonomous systems more transparent.
DIN SPEC 92001-1

Published

General, Developers

Quality, Safety

Details

Standards Details

Title
DIN SPEC 92001-1

Domain

Subject matter

Standard type
Standard

Status
Published

Publisher
DIN

Audience
General, Developers

AI Aspect
Quality, Safety

Buy Standard

Summary
This DIN SPEC in accordance with the PAS procedure has been drawn up by a DIN SPEC (PAS)-consortium set up on a temporary basis. This DIN SPEC (PAS) has been developed and approved by the authors named in the foreword. This DIN SPEC is divided into two parts. Part 1, DIN SPEC 92001-1, provides a general quality metal model for artificial intelligence (AI) that will primarily describe the most important aspects of AI quality. The AI quality metal model features quality characteristics, most notably performance & functionality, robustness and comprehensibility. This document also deals with the AI module as part of a software system and assesses the risks associated with this module; it also describes a suitable software life cycle approach. The present AI life cycle approach leans heavily on the International Standard on system and software development ISO/IEC/IEEE 12207:2017. The aim of this series of DIN SPECs is to enable the safe, transparent development and use of AI modules. To achieve this, the DIN SPEC describes a number of quality requirements that are structured using an AI quality meta model. The DIN SPEC series applies to all phases in the life cycle of an AI module (i.e. its conceptualization, development, use, operation and termination) and takes into account a number of different life cycle processes. AI technologies are used for a diverse range of applications, which is why this DIN SPEC series is aimed at companies across all sectors. © 2022 Beuth Verlag GmbH
DIN SPEC 92001-2

Published

General, Developer, Provider

Robustness, System lifecycle, System quality

Details

Standards Details

Title
DIN SPEC 92001-2

Domain

Subject matter

Standard type
National Standard

Status
Published

Publisher
DIN

Audience
General, Developer, Provider

AI Aspect
Robustness, System lifecycle, System quality

Buy Standard

Summary
Artificial Intelligence – Life Cycle Processes and Quality Requirements – Part 2: Robustness. This DIN SPEC in accordance with the PAS procedure has been drawn up by a DIN SPEC (PAS)-consortium set up on a temporary basis. This DIN SPEC (PAS) has been developed and approved by the authors named in the foreword. The aim of this series of DIN SPECs is to enable the safe and transparent development and use of AI modules. To achieve this, the DIN SPEC describes a number of AI quality requirements which are structured using an AI quality meta model. The DIN SPEC series applies to all phases of the life cycle of an AI module (i.e. its conceptualization, development, use, operation and termination) and takes into account a number of different life cycle processes. AI technologies are used for a diverse range of applications, which is why this DIN SPEC series is aimed at companies across all sectors. Part 2 of the DIN SPEC 92001 series describes the quality requirements specific to AI robustness. These quality requirements are structured using the specified AI quality meta model (DIN SPEC 92001-1). DIN SPEC 92001-2 explains various mathematical AI robustness requirements, specifically regarding Adversarial Robustness (i.e. AI attacks with mathematically optimized perturbations leading to model failure) and Corruption Robustness (i.e. model sensitivities to naturally occurring noise / data outliers leading to model failure). © 2022 Beuth Verlag GmbH
IEEE 2801

Published

General, Providers, Developers

All

Details

Standards Details

Title
IEEE 2801

Domain

Subject matter

Standard type
Standard

Status
Published

Publisher
IEEE

Audience
General, Providers, Developers

AI Aspect
All

Buy Standard

Summary
IEEE Recommended Practice for the Quality Management of Datasets for Medical Artificial Intelligence Promoted in this recommended practice are quality management activities for datasets used for artificial intelligence medical devices (AIMD). The document highlights quality objectives for organizations responsible for datasets. The document describes control of records during the lifecycle of datasets, including but not limited to data collection, annotation, transfer, utilization, storage, maintenance, updates, retirement, and other activities. The document emphasizes special consideration for the dataset quality management system, including but not limited to responsibility management, resource management, dataset realization, and quality control. © Copyright 2023 IEEE – All rights reserved.
IEEE 7000

Published

General, Developers

Accountability, Bias and discrimination, Explainability and transparency, Privacy, Project management, Risk management, Stakeholder engagement and communication, Sustainability

Details

Standards Details

Title
IEEE 7000

Domain

Subject matter

Standard type
Standard

Status
Published

Publisher
IEEE

Audience
General, Developers

AI Aspect
Accountability, Bias and discrimination, Explainability and transparency, Privacy, Project management, Risk management, Stakeholder engagement and communication, Sustainability

Buy Standard

Summary
Abstract A set of processes by which organizations can include consideration of ethical values throughout the stages of concept exploration and development is established by this standard. Management and engineering in transparent communication with selected stakeholders for ethical values elicitation and prioritization is supported by this standard, involving traceability of ethical values through an operational concept, value propositions, and value dispositions in the system design. Processes that provide for traceability of ethical values in the concept of operations, ethical requirements, and ethical risk-based design are described in the standard. All sizes and types of organizations using their own life cycle models are relevant to this standard. © Copyright 2022 IEEE Scope The standard establishes a set of processes by which engineers and technologists can include consideration of ethical values throughout the stages of concept exploration and development, which encompass system initiation, analysis, and design. This standard provides engineers and technologists with an implementable process aligning innovation management processes, system design approaches, and software engineering methods to help address ethical concerns or risks during system design. IEEE Std 7 000(TM) does not give specific guidance on the design of algorithms to apply ethical values such as fairness and privacy. © IEEE 2022 All rights reserved. Purpose The goal of this standard is to enable organizations to design systems with explicit consideration of individual and societal ethical values, such as transparency, sustainability, privacy, fairness, and accountability, as well as values typically considered in system engineering, such as efficiency and effectiveness. Projects conforming to IEEE Std 7000 balance management commitments for time and budget constraints with the long-term values of social responsiveness and accountability. To enable this, the commitment of top executives to establish and uphold organizational values is important. © IEEE 2022 All rights reserved.
NIST AI RMF 1.0

Published

General, Developers, Providers

Process, Management and Governance

Details

Standards Details

Title
NIST AI RMF 1.0

Domain

Subject matter

Standard type
Voluntary Framework

Status
Published

Publisher
NIST - National Institute of Standards and Technology

Audience
General, Developers, Providers

AI Aspect
Process, Management and Governance

Buy Standard

Summary
In collaboration with the private and public sectors, NIST has developed a framework to better manage risks to individuals, organizations, and society associated with artificial intelligence (AI). The NIST AI Risk Management Framework (AI RMF) is intended for voluntary use and to improve the ability to incorporate trustworthiness considerations into the design, development, use, and evaluation of AI products, services, and systems. Released on January 26, 2023, the Framework was developed through a consensus-driven, open, transparent, and collaborative process that included a Request for Information, several draft versions for public comments, multiple workshops, and other opportunities to provide input. It is intended to build on, align with, and support AI risk management efforts by others. A companion NIST AI RMF Playbook also has been published by NIST along with an AI RMF Roadmap, AI RMF Crosswalk, and various Perspectives. In addition, NIST is making available a video explainer about the AI RMF. To view public comments received on the previous drafts of the AI RMF and Requests for Information, see the AI RMF Development page. NIST Standard Reference Data (SRD); ©Copyright 2023 by the U.S. Secretary of Commerce on behalf of the United States of America. All rights reserved.
ISO/IEC Guide 51:2014

Published

Standardization

Safety

Details

Standards Details

Title
ISO/IEC Guide 51:2014

Domain

Subject matter

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
Standardization

AI Aspect
Safety

Buy Standard

Summary
ISO/IEC Guide 51:2014 provides requirements and recommendations for the drafters of standards for the inclusion of safety aspects in standards. ISO/IEC Guide 51:2014 is applicable to any safety aspect related to people, property or the environment, or to a combination of these.
ISO/IEC 27001:2022

Published

Developers

Cybersecurity

Details

Standards Details

Title
ISO/IEC 27001:2022

Domain

Subject matter

Standard type
Standard

Status
Published

Publisher
ISO/IEC

Audience
Developers

AI Aspect
Cybersecurity

Buy Standard

Summary
ISO/IEC 27001 is the world's best-known standard for information security management systems (ISMS). It defines requirements an ISMS must meet. The ISO/IEC 27001 standard provides companies of any size and from all sectors of activity with guidance for establishing, implementing, maintaining and continually improving an information security management system. Conformity with ISO/IEC 27001 means that an organization or business has put in place a system to manage risks related to the security of data owned or handled by the company, and that this system respects all the best practices and principles enshrined in this International Standard. Why is ISO/IEC 27001 important? With cyber-crime on the rise and new threats constantly emerging, it can seem difficult or even impossible to manage cyber-risks. ISO/IEC 27001 helps organizations become risk-aware and proactively identify and address weaknesses. ISO/IEC 27001 promotes a holistic approach to information security: vetting people, policies and technology. An information security management system implemented according to this standard is a tool for risk management, cyber-resilience and operational excellence.

Term

Definition

Source
Accessibility

Extent to which products, systems, services, environments and facilities can be used by people from a population with the widest range of user needs, characteristics and capabilities to achieve identified goals in identified contexts of use (which includes direct use or use supported by assistive technologies).

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Accountability

the ability to explain and justify one's actions.

EU AI Act
Adaptive learning

An adaptive AI is a system that changes its behaviour while in use. Adaptation may entail a change in the weights of the model or a change in the internal structure of the model itself. The new behaviour of the adapted system may produce different results than the previous system for the same inputs.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Adversarial machine learning (adversarial attack)

A practice concerned with the design of ML algorithms that can resist security challenges, the study of the capabilities of attackers, and the understanding of attack consequences. Inputs in adversarial ML are purposely designed to make a mistake in its predictions despite resembling a valid input to a human.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
(AI) accuracy

Closeness of computations or estimates to the exact or true values that the statistics were intended to measure. The goal of an AI model is to learn patterns that generalise well for unseen data. It is important to check if a trained AI model is performing well on unseen examples that have not been used for training the model. To do this, the model is used to predict the answer on the test dataset and then the predicted target is compared to the actual answer. The concept of accuracy is used to evaluate the predictive capability of the AI model. Informally, accuracy is the fraction of predictions the model got right. A number of metrics are used in machine learning (ML) to measure the predictive accuracy of a model. The choice of the accuracy metric to be used depends on the ML task.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
AI (or algorithmic) bias

Harmful AI bias describes systematic and repeatable errors in AI systems that create unfair outcomes, such as placing. privileged groups at systematic advantage and unprivileged groups at systematic disadvantage. Different types of bias can emerge and interact due to many factors, including but not limited to, human or system decisions and processes across the AI lifecycle. Bias can be present in AI systems resulting from pre-existing cultural, social, or institutional expectations; because of technical limitations of their design; by being used in unanticipated contexts; or by non-representative design specifications.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
AI regulatory sandbox

a controlled environment set by authorities for testing innovative AI systems under supervision for a limited time.

EU AI Act
AI system

software that can, for a given set of input data, generate outputs such as predictions, decisions, or recommendations without being explicitly programmed for each specific input-output mapping.

EU AI Act
Algorithm

An algorithm consists of a set of step-by-step instructions to solve a problem (e.g., not including data). The algorithm can be abstract and implemented in different programming languages and software libraries.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Artificial Intelligence Office

a part of the European Commission that oversees AI systems, models, and governance.

EU AI Act
Attack

Action targeting a learning system to cause malfunction.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Auditability of an AI system

Auditability refers to the ability of an AI system to undergo the assessment of the system’s algorithms, data and design processes. This does not necessarily imply that information about business models and Intellectual Property related to the AI system must always be openly available. Ensuring traceability and logging mechanisms from the early design phase of the AI system can help enable the system's auditability.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Autonomy (autonomous AI system)

Systems that maintain a set of intelligence-based capabilities to respond to situations that were not pre-programmed or anticipated (i.e., decision-based responses) prior to system deployment. Autonomous systems have a degree of self-government and self-directed behaviour (with the human’s proxy for decisions).

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Big data

An all-encompassing term for large, complex digital data sets that need equally complex technological means to be stored, analysed, managed and processed with substantial computing power. Datasets are sometimes linked together to see how patterns in one domain affect other areas. Data can be structured into fixed fields or unstructured as free-flowing information. The analysis of big datasets, often using AI, can reveal patterns, trends, or underlying relationships that were not previously apparent to researchers.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Biometric categorisation system

an AI system that groups people based on their biometric data, unless it's a necessary part of another service.

EU AI Act
Biometric data

Personal data from physical, physiological, or behavioural traits used to identify a person, like fingerprints or facial recognition.

EU AI Act
Biometric identification

Using AI to recognize a person's unique physical or behavioural traits to confirm who they are by comparing their data to a database.

EU AI Act
Biometric verification

confirm someone's identity by comparing their biometric data to previously stored data (one-to-one verification, including authentication).

EU AI Act
CE marking of conformity (CE marking)

A symbol that shows an AI system meets specific European Union regulations and standards for being sold.

EU AI Act
Chatbot (conversational bot)

A computer program designed to simulate conversation with a human user, usually over the internet; especially one used to provide information or assistance to the user as part of an automated service.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Classification

A classification system is a set of “boxes” into which things are sorted. Classifications are consistent, have unique classificatory principles, and are mutually exclusive. In AI design, when the output is one of a finite set of values (such as sunny, cloudy or rainy), the learning problem is called classification, and is called Boolean or binary classification if there are only two values.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Classifier

A model that predicts (or assigns) class labels to data input.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Data poisoning

A type of security attack where malicious users inject false training data with the aim of corrupting the learned model, thus making the AI system learn something that it should not learn.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Data protection and transparency

the obligations on providers and users of AI systems to protect personal data and to be transparent about the operation of their systems.

EU AI Act
Deep learning

A subset of machine learning based on artificial neural networks that employs statistics to spot underlying trends or data patterns and applies that knowledge to other layers of analysis. Some have labelled this as a way to “learn by example” and as a technique that “perform[s] classification tasks directly from images, text, or sound” and then applies that knowledge independently.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Deployer

an organisation using an AI system under their authority except for personal non-professional use

EU AI Act
Differential privacy

Differential privacy is a method for measuring how much information the output of a computation reveals about an individual. It produces data analysis outcomes that are nearly equally likely, whether any individual is, or is not, included in the dataset. Its goal is to obscure the presence or absence of any individual (in a database), or small groups of individuals, while at the same time preserving statistical utility.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Discrimination

Unequal treatment of a person based on belonging to a category rather than on individual merit. Discrimination can be a result of societal, institutional and implicitly held individual biases or attitudes that get captured in processes across the AI lifecycle, including by AI actors and organisations, or represented in the data underlying AI systems. Discrimination biases can also emerge due to technical limitations in hardware or software, or the use of an AI system that, due to its context of application, does not treat all groups equally. Discriminatory biases can also emerge in the very context in which the AI system is used. As many forms of biases are systemic and implicit, they are not easily controlled or mitigated and require specific governance and other similar approaches"

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Distributor

an organisation, other than the provider or importer, that makes an AI system available in the EU

EU AI Act
Downstream provider

a provider of an AI system, including a general- purpose AI system, which integrates an AI model, regardless of whether the model is provided by themselves and vertically integrated or provided by another entity based on contractual relations.

EU AI Act
EU Declaration of Conformity

a document that states that an AI system has been assessed and found to meet the requirements of the EU AI Act.

EU AI Act
Evaluation

Systematic determination of the extent to which an entity meets its specified criteria.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Federated learning

Federated learning is a machine learning model which addresses the problem of data governance and privacy by training algorithms collaboratively without transferring the data to another location. Each federated device shares its local model parameters instead of sharing the whole dataset used to train it and the federated learning topology defines the way parameters are shared.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Floating-point operation

any mathematical operation or assignment involving floating-point numbers, which are a subset of the real numbers typically represented on computers by an integer of fixed precision scaled by an integer exponent of a fixed base.

EU AI Act
Fundamental rights and values

the rights and values enshrined in the Charter of Fundamental Rights of the European Union.

EU AI Act
General purpose AI model

an AI model that can do many different tasks well, regardless of how it's used or integrated into other systems or applications.

EU AI Act
General purpose AI system

an AI system which is based on a general purpose AI model, that has the capability to serve a variety of purposes, both for direct use as well as for integration in other AI systems;

EU AI Act
Generative adversarial network (GAN)

Generative Adversarial Networks, or GANs for short, are an approach to generative modelling using deep learning methods, such as convolutional neural networks. Generative modelling is an unsupervised learning task in machine learning that involves automatically discovering and learning the regularities or patterns in input data in such a way that the model can be used to generate or output new examples that plausibly could have been drawn from the original dataset.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Harm

any negative impact on individuals, society, or the environment.

EU AI Act
High-risk AI system

any AI system that is likely to cause or increase harm to individuals, society, or the environment.

EU AI Act
Human oversight

the requirement for high-risk AI systems to have mechanisms in place to allow humans to override the decisions of the systems and to monitor their operation.

EU AI Act
Human values for AI

Values are idealised qualities or conditions in the world that people find good. AI systems are not value-neutral. The incorporation of human values into AI systems requires that we identify whether, how and what we want AI to mean in our societies. It implies deciding on ethical principles, governance policies, incentives, and regulations. And it also implies that we are aware of differences in interests and aims behind AI systems developed by others according to other cultures and principles.functioning of biological neurons in the nervous system, most neural networks are used in artificial intelligence as realisations of the connectionist model.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Importer

an organisation importing and selling into the EU market an AI system that bears the name or trademark of an organisation located outside the EU

EU AI Act
Input data

data provided to or directly acquired by an AI system on the basis of which the system produces an output.

EU AI Act
Intended purpose

the use for which an AI system is intended by the provider, including the specific context and conditions of use, as specified in the information supplied by the provider in the instructions for use, promotional or sales materials and statements, as well as in the technical documentation.

EU AI Act
Law enforcement

activities by authorities to prevent, investigate, or prosecute crimes, or to protect public safety.

EU AI Act
Non-high-risk AI system

any AI system that is not likely to cause or increase harm to individuals, society, or the environment.

EU AI Act
Non-personal data

data other than personal data

EU AI Act
Notified Body

an independent organization that is authorized by the EU Commission to assess the conformity of AI systems.

EU AI Act
Operator

the provider, the product manufacturer, the deployer, the authorised representative, the importer or the distributor

EU AI Act
Personal data

Information that can identify a person, as defined by article 4 of the GDPR

EU AI Act
Post-market monitoring

the obligation on providers of AI systems to monitor the performance of their systems and to take corrective action if necessary.

EU AI Act
Post-market monitoring system

Activities by AI system providers to collect feedback from users to see if any improvements or fixes are needed.

EU AI Act
Post remote biometric identification system

a system that identifies people but not in real-time, so there's some delay.

EU AI Act
Prior conformity assessment

the process of having an AI system assessed by an independent body to verify that it complies with the requirements of the EU AI Act.

EU AI Act
Provider

an organisation developing or commissioning the development of an AI system and selling or putting into service under their own name or trademark

EU AI Act
Real-time remote biometric identification system

a system that identifies people almost immediately without delay.

EU AI Act
Reasonably foreseeable misuse

the use of an AI system in a way that is not in accordance with its intended purpose, but which may result from reasonably foreseeable human behaviour or interaction with other systems, including other AI systems.

EU AI Act
Remote biometric identification system

an AI system that identifies people from a distance without their active participation by comparing their biometric data to a database.

EU AI Act
Safety

the state of not being at risk of harm.

EU AI Act
Scalability

The ability to increase or decrease the computational resources required to execute a varying volume of tasks, processes, or services.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Socio-technical system

Technology is always part of society, just like society is always part of technology. This also means that one cannot understand one without the other. Technology is not only design and material appearance but also sociotechnical; that is, a complex process constituted by diverse social, political, economic, cultural and technological factors.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Standard

Standards are a set of institutionalised agreed upon-rules for the production of (textual or material) objects. They are released by international organisations and ensure quality and safety and set product or services’ specifications. Standards are the result of negotiations among various stakeholders and are institutionalised and thus difficult to change.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Substantial modification

A change to an AI system after it's been released that was not planned in the initial conformity assessment by the provide and affects how well it meets certain requirements or changes its intended use.

EU AI Act
Technical interoperability

The ability of software or hardware systems or components to operate together successfully with minimal effort by an end user.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Test

Technical operation to determine one or more characteristics of or to evaluate the performance of a given product, material, equipment, organism, physical phenomenon, process or service according to a specified procedure.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Test and Evaluation, Verification and Validation (TEVV)

A framework for assessing, incorporating methods and metrics to determine that a technology or system satisfactorily meets its design specifications and requirements, and that it is sufficient for its intended use.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Testing data

data used for providing an independent evaluation of the AI system to confirm the expected performance

EU AI Act
Training data

data used for training an AI system through fitting its learnable parameters

EU AI Act
Transparency

the ability to be easily understood.

EU AI Act
Unacceptable AI system

any AI system that is likely to pose an unacceptable risk to human safety or fundamental rights and values.

EU AI Act
Validation

Confirmation by examination and provision of objective evidence that the particular requirements for a specific intended use are fulfilled.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Validation data

means data used for providing an evaluation of the trained AI system and for tuning its non-learnable parameters and its learning process

EU AI Act
Value sensitive design (values-by- design or ethics- by-design)

A theoretically grounded approach to the design of technology that accounts for human values in a principled and systematic manner throughout the design process.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence
Verification

Provides evidence that the system or system element performs its intended functions and meets all performance requirements listed in the system performance specification.

EU-U.S. Terminology and Taxonomy for Artificial Intelligence

What is AI Assurance?

AI Assurance refers to the process and methodologies used to ensure that AI systems operate safely, ethically, and in compliance with relevant regulations and standards.

How can I start learning about AI?

Go to our Knowledge Hub of AI Assurance and begin with our "Introduction to AI" section, which covers fundamental concepts such as the agreed definition and properties of AI.

What is AI ethics?

AI ethics involves the study and evaluation of ethical problems associated with AI and automated systems, including issues of bias, privacy, accountability, and the impact on society. In our knowledge hub, we dive into specific concepts of Trustworthy AI.

What are the latest regulatory developments in the US and EU regarding AI?

Our libraries are continuously updated with the latest regulatory developments in both the US and EU. Check the Regulations Library for the most current information.

How do US and EU regulations on AI differ?

The US focuses on sector-specific guidelines and fostering innovation, while the EU aims for a comprehensive, risk-based regulatory framework emphasizing ethical standards and societal values.

What standards and normative frameworks exist for AI?

Explore our library of standards and normative frameworks to understand the international and regional guidelines established for AI development and deployment.

How can standards help in achieving AI assurance?

Standards provide a baseline for quality, safety, and ethics in AI systems, helping organizations align their AI initiatives with best practices and regulatory requirements.

Where can I find regulations and standards related to AI?

Our Libraries section offers comprehensive access to regulations, standards, audit catalogues, and toolkits relevant to AI.

Can I contribute to the libraries?

Yes, contributions are welcome! Please contact our team for guidelines on contributing to our collaborative environment. You can add your inputs in this form.

How can the AI trust agent help me?

Our AI trust agent can provide instant answers to your questions about AI assurance, guide you through the environment, and suggest resources related to your queries. Note that the agent is currently in its beta version and steady improvements will be make over the coming weeks.

What should I do if the chatbot can't answer my question?

If the chatbot is unable to provide an answer it will suggest additional resources for further research.

How often is the content in the AI Trust Hub updated?

We strive to keep our content current with the latest developments in AI assurance. The environment is reviewed and updated regularly.

What is the purpose of the World Map?

The World Map aims to easily illustrate and help you visualize the state of AI governance across various countries. It is color-coded to distinguish between countries with specific AI policies (dark blue) and those subject to general global policies (light blue).

What kind of content can I expect in the News Feed?

We will be regularly populating and updating the News Feed with content related to the latest AI policy developments around the world, as well as with information about our newest features and services provided by Nemko Digital.

What is the AI Governance self-assessment tool?

The AI Governance self-assessment tool allows you to perform a preliminary evaluation of the extent of AI governance in your organization, identifying opportunities for improvement in one or more areas. It is compatible with Nemko’s AI Trust services, which can help you increase your self-assessment score and achieve effective and compliant AI governance within your organization.

Book Your Free Consultation Call

Ready to elevate your AI product’s trustworthiness and compliance?
Don’t wait - connect with our experts for a free 15-minute consultation call to discuss how our trusted services can help your business thrive in today's competitive landscape.

Schedule Your 15-Minute Call Now

Webinars

Enterprise AI Agents: Productivity vs Compliance Risks

Why Trust in AI Needs a Global Framework

ISO/IEC 42005: AI Impact Assessment Framework Guide

AI Medical Software Compliance: Navigating EU Regulations

AI Maturity Readiness: The 8-Dimension Approach to Success

Dubai AI Week 2025: Reflections on UAE's Accelerating AI Journey | Nemko Digital

ISO 42001 Certification: The Key to AI Regulatory Compliance

AI Maturity Governance: Why It Matters for Your Business

The UAE’s Visionary Approach to AI: Innovation with Ethics at the Core

Global AI trust certification: AI Trust Mark unveiled - Nemko Digital

Navigating the EU AI Act in 2025: Key Actions and Compliance Strategies

Navigating the EU AI Act: A Strategic Approach to AI Product Risk Evaluation

Fundamental Rights Impact Assessments (FRIAs) under the EU AI Act: What You Need to Know

How can inaccurate AI harm a business?

The AI Lifecycle

General Product Safety Regulation (GPSR)

Cybersecurity Landscape in the Field of AI

A Pivotal Year for AI Governance and the Road Ahead

Understanding the First-Ever International AI Treaty: A Legal Milestone

Navigating the World of AI Assurance: Nemko's Strategic Move

Regulating Artificial Intelligence

Ensuring a fair future: The crucial role of ethics in AI development

A Quick Dive into the EU AI Act: Key Insights & Implications

Navigating the Future of Artificial Intelligence with ISO 42001: A Guide for Businesses

Keeping AI in Check: The Critical Role of Human Agency and Oversight

The Foundations of AI Safety: Exploring Technical Robustness

Transparency in AI as a Competitive Advantage

Mastering AI Privacy and Data Governance

Diversity, Non-Discrimination, and Fairness in AI Systems

Environmental and Societal Wellbeing: A Key Requirement for Trustworthy AI

Webinars Library

Digital Portfolio

Newsletters

Search

Search

Search

Book Your Free Consultation Call

Webinars

Enterprise AI Agents: Productivity vs Compliance Risks

Why Trust in AI Needs a Global Framework

ISO/IEC 42005: AI Impact Assessment Framework Guide

AI Medical Software Compliance: Navigating EU Regulations

AI Maturity Readiness: The 8-Dimension Approach to Success

Dubai AI Week 2025: Reflections on UAE's Accelerating AI Journey | Nemko Digital

ISO 42001 Certification: The Key to AI Regulatory Compliance

AI Maturity Governance: Why It Matters for Your Business

The UAE’s Visionary Approach to AI: Innovation with Ethics at the Core

Global AI trust certification: AI Trust Mark unveiled - Nemko Digital

Navigating the EU AI Act in 2025: Key Actions and Compliance Strategies

Navigating the EU AI Act: A Strategic Approach to AI Product Risk Evaluation

Fundamental Rights Impact Assessments (FRIAs) under the EU AI Act: What You Need to Know

How can inaccurate AI harm a business?

The AI Lifecycle

General Product Safety Regulation (GPSR)

Cybersecurity Landscape in the Field of AI

A Pivotal Year for AI Governance and the Road Ahead

Understanding the First-Ever International AI Treaty: A Legal Milestone

Navigating the World of AI Assurance: Nemko's Strategic Move

Regulating Artificial Intelligence

Ensuring a fair future: The crucial role of ethics in AI development

A Quick Dive into the EU AI Act: Key Insights & Implications

Navigating the Future of Artificial Intelligence with ISO 42001: A Guide for Businesses

Keeping AI in Check: The Critical Role of Human Agency and Oversight

The Foundations of AI Safety: Exploring Technical Robustness

Transparency in AI as a Competitive Advantage

Mastering AI Privacy and Data Governance

Diversity, Non-Discrimination, and Fairness in AI Systems

Environmental and Societal Wellbeing: A Key Requirement for Trustworthy AI

Webinars Library

Digital Portfolio

Nemko Digital Service Brochure

Newsletters

Newsletter February 2025

Newsletter January 2025

Search

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Regulations Details

Search

Standards Details

Standards Details