AI Adoption in Conformity Assessment

A new survey from the AIQI Consortium found that 21% of conformity assessment bodies (CABs) are currently using AI, and another 15% are in active development. Combined, that 36% is significantly lower than the 88% adoption rate reported across other industries for using AI in at least one business function. The gap is not a failure. It reflects something specific to this sector.

How AI Is (and Isn’t) Used in Conformity Assessment

Among the 21% of CABs that have adopted AI, most are using it in supporting functions such as communications, training, marketing, and scheduling. Eighty-four percent of CABs using AI apply it to at least some supporting processes, and 54% use it exclusively. Their core conformity assessment activities tell a different story. Selection, determination, review, and decision-making have seen much more limited uptake. Only 16% of AI-using CABs apply AI directly to conformity assessment processes.

This distribution makes sense. Off-the-shelf tools can be deployed quickly, without customization or significant change-management. Using AI in an audit, an inspection, or a certification decision is a different proposition. It requires systems tailored to specific technical domains, careful thought about where professional judgment begins and ends, and clarity from accreditation bodies about what acceptable use actually looks like.

That last point is where the data gets most relevant for accredited organizations.

AI in Accreditation: Why Clear Guidance Matters for CABs

The AIQI Consortium report identifies the absence of clear expectations from accreditation bodies as one factor limiting AI adoption in core conformity assessment activities. CABs operating under accreditation are accustomed to working within defined requirements. When those requirements don’t yet address a rapidly evolving area, uncertainty tends to produce inaction rather than experimentation.

This isn’t unique to AI. It’s the same dynamic that plays out whenever new technology intersects with technically rigorous, third-party processes.

ANAB has begun addressing this directly. We have established a policy position on how assessors may use AI tools during accreditation assessments. The standards and accreditation landscape around AI is still developing, and our guidance will continue to evolve alongside it.

AI Tools in Conformity Assessment

No single AI type dominates among CABs using or developing AI. Machine learning and predictive analytics each come in at 12%, large language models at 8%, and agentic AI at just 4%. The report notes that many “other” responses cited brand names rather than AI categories, which suggests limited familiarity with the underlying technology types.

That observation matters. CABs considering AI adoption in technical processes need working fluency in the technology, not just familiarity with the tools already in use. Understanding what machine learning does differently than a decision support system, or what agentic AI introduces in terms of autonomy and risk, is relevant when making decisions about implementation in accredited activities.

The sector is early in this. Organizations that take time now to understand the technology, track the emerging standards landscape, and think carefully about where AI adds genuine value to their processes will be better positioned as expectations solidify.

ANAB’s AI Policies and Future Plans

ANAB has established AI policies for both internal staff and assessors, launched assessor training on AI for information systems, and formed an internal AI task force to identify where AI can be applied across our operations. Looking ahead, we are exploring the development of AI agents to support accreditation processes and activities. We expect to share more as that work progresses.

For CABs thinking through their own AI programs, the questions the AIQI data raises are worth taking seriously now: where does AI fit in your processes, what does your accreditation body expect, and where does professional judgment remain the standard?

The full AIQI Consortium report is available at aiqi.org.