AI & Cyber Testing and Assurance

Independent assurance provides structured, evidence-based evaluation of an AI system against defined criteria. It can help organizations demonstrate trustworthiness to customers, partners, procurement teams, and other stakeholders, while identifying opportunities for improvement. For organizations bringing AI-enabled products or services to market, this can support customer confidence and commercial differentiation.

The AI Trust Mark is an independent assessment and certification scheme for AI systems against defined trustworthiness criteria. ISO/IEC 42001, by contrast, focuses on an organization's AI management system. The two approaches are complementary: organizational governance establishes the processes for managing AI, while system-level assurance provides evidence about the characteristics of a specific AI system.

The AI Trust Mark can provide valuable evidence relevant to AI trustworthiness and certain regulatory requirements, depending on the assessment scope. However, it should not be treated as an automatic substitute for conformity assessment or other legal obligations under the EU AI Act. We help organizations understand how the assessment relates to their specific regulatory responsibilities and where additional activities may be necessary.

The assessment scope depends on the AI system's intended use, characteristics, risk profile, and the applicable scheme requirements. Relevant areas may include robustness, fairness, transparency, security, and human oversight. We begin by clarifying the scope and reviewing available documentation and evidence, then identify any gaps that need to be addressed before formal assessment.

The assessment is performed through the applicable independent assurance and certification arrangements. The process typically begins with defining the system and assessment scope, followed by a readiness review, evidence evaluation, and required testing or assessment activities. Where advisory and certification services are both involved, appropriate independence and impartiality requirements must be maintained. The exact arrangements are confirmed as part of the engagement.

Trustworthy AI requires ongoing governance, monitoring, and change management as systems, data, and operating conditions evolve. We help organizations establish processes to assess material changes, maintain relevant evidence, and identify when additional evaluation may be needed. Any surveillance, reassessment, or renewal requirements are determined by the applicable certification scheme.