ISO/IEC TR 24028:2020
This standard establishes an approach for measuring the trustworthiness of AI systems.
ISO/IEC TR 24028 surveys the existing approaches to examining trustworthiness in AI systems and provides a level of quality to expect from these approaches. With the help of this standard, organizations can maximize the efficacy of their trust-building strategies.
How does ISO/IEC TR 24028 respond to our increasing dependence on artificial intelligence?
As more and more tasks in various industries are performed by AI, it is of utmost importance to our society to monitor the trustworthiness of AI systems. Replacing human intelligence with automated tools in the development of new products and services is an excellent opportunity for efficiency of both time and cost. However, passing responsibility onto these technologies means that organizations must be rigorous at maintaining a high standard of accountability for the systems they employ. ISO/IEC TR 24028 is a key tool for directing your systems towards being part of a credible name for AI.
The trustworthiness of AI systems can be measured by two key qualities:
- Transparency in regard to data quality, storage, and handling; design choices and procedures; parties involved in the development of the system. Thorough documentation is essential to a trustworthy system. Organizations should perform frequent audits in order to keep stakeholders informed on internal decisions and their outcomes.
- Risk preparedness in response to possible pitfalls in the AI system. This includes not only a robust plan for dealing with possible security issues, but also mitigation techniques for the impact that the system could have on users, societies, and the environment. A responsible system is aware of its possible impact on humans and other technologies and emphasizes resiliency as a consequence.
ISO/IEC TR 24028 describes the approaches that are most effective in creating high transparency and high preparedness. Additionally, it outlines how to assess your system for critical gaps in security, accuracy, and reliability.
This standard is relevant for not only the merchants of these technologies, but also for developers, machine learning engineers, data scientists, social researchers, and localized regulatory bodies. Adopting ISO/IEC TR 24028 helps you to create a more ethical and more efficient AI system using international-grade guidance. It allows for faster innovation within your organization as it simplifies preparedness for regulatory compliance. With trust in mind, let’s build a credible name for AI