The assurance economy: ISO 42001 and the new profession

The growing demand for trustworthy artificial intelligence (AI) systems is driving the emergence of a new professional field: AI assurance. As organisations seek to demonstrate the reliability and safety of their AI systems, a structured approach to evaluation and certification is becoming essential. This evolution is supported by both public and private sector initiatives aimed at establishing standards and competencies.

Market Growth and Investment in AI Trustworthiness

A market is forming around proving AI trustworthiness, with the UK Department for Science, Innovation and Technology (DSIT) estimating that approximately 12,500 people are currently engaged in AI assurance-related activities. DSIT’s roadmap projects that this market will grow to 18.8 billion pounds by 2035. This growth is backed by a £11 million innovation fund and a broader strategy to professionalise the field.

Standardisation and Certification Frameworks

The management-system anchor for this professionalisation is ISO/IEC 42001, an international standard for AI governance. On 15 January 2026, the British Standards Institution (BSI) received the first UKAS accreditation to certify AI management systems. The certification body NQA followed shortly after, with LRQA also gaining accreditation. Darktrace became an early adopter, certifying its AI management systems under this framework. The EU AI Act’s high-risk obligations, effective from 2 August 2026, are expected to further accelerate demand for such certifications.

Professionalisation and Institutional Frameworks

The AI assurance profession is being deliberately structured through a collaborative effort involving public and private stakeholders. The AI Assurance Consortium, chaired by the British Computer Society (BCS), is working alongside techUK, which leads the industry advisory group. These initiatives are complemented by the development of a skills and competencies framework and a code of ethics, both of which are expected to be implemented in the near future.

Government Support for Smaller Organisations

Recognising that smaller organisations may lack resources for full compliance, the UK government has introduced tools to support AI governance at all scales. The AI Management Essentials programme offers a self-assessment tool that helps smaller entities evaluate their AI practices. Additionally, the Responsible Technology Adoption Unit provides applied guidance, such as the publication on responsible AI in recruitment, to assist with practical implementation.

Limitations of Certification and the Role of Training

Certification under frameworks like ISO 42001 does not replace the need for technical evaluation. A case study from the evaluation course in this series demonstrates that a weak system can still be certified if the underlying technical evaluation is insufficient. This highlights the importance of critical reading and understanding that a certificate alone does not guarantee robustness. Training in evaluation methodologies is therefore essential, particularly given the anticipated scale of the assurance market. Tesseract Academy’s course offerings and the instructor’s published methods are tailored to meet the growing demand for skilled evaluators in this sector.

What to take away

The assurance economy around AI is rapidly developing, with certification standards like ISO/IEC 42001 playing a central role. However, certification must be underpinned by rigorous technical evaluation. The expansion of this field requires both institutional frameworks and skilled professionals, creating demand for structured training and education.

Reference

Sources and further reading

• First UKAS ISO/IEC 42001 accreditation
• techUK on the AI Assurance Consortium