ISO/IEC 17024:2026: What Changes When AI Enters the Certification Workflow?

The 2026 revision of ISO/IEC 17024 comes as certification programs rely more heavily on digital delivery and automated tools. Published on March 31, 2026, the third edition replaces the 2012 standard and adds explicit requirements governing the use of AI in certification processes. A system may work as designed and still leave an important question unanswered: can the certification body explain how it shaped the assessment outcome?

What the revision puts in focus

AI is no longer a side issue

ISO's overview highlights a tailored definition of AI and dedicated requirements for its use in certification processes. It points to monitoring and validating AI-generated outcomes, maintaining human oversight, demonstrating validity, reliability and fairness, and ensuring that people overseeing AI tools are competent to do so. ANAB describes the new provisions in terms of transparency, oversight, validation, and disclosure.

That matters because AI is already present across credentialing workflows, from test development and scoring support to remote proctoring and operational monitoring. The question is no longer simply whether AI can improve efficiency. Certification bodies need to understand where it is used, what role it plays, and how people remain accountable for the process around it.

A more harmonized structure

The 2026 edition also updates terminology, aligns the standard with common CASCO elements, and adds an informative Annex B showing how certification activities, processes, assessments, applications, examinations, and schemes relate. The result is a clearer view of where individual activities sit within the wider certification system.

Together, these updates support the standard’s broader objective: greater consistency, comparability, and reliability across borders and sectors.

What should certification bodies consider?

The revision should not be reduced to a technology checklist. It does, however, give certification leaders a useful reason to examine how modern assessment operations are governed. Three questions are worth asking.

1. Where is AI used, and how is human oversight maintained?

If AI supports proctoring, scoring, content development, or other assessment activities, can the organization clearly explain where it enters the workflow? Can it show how outputs are reviewed or validated, and where a person remains responsible for judgment, escalation, or decision-making? Mapping that boundary is a practical first step toward making AI use visible and governable.

2. Can the program reconstruct what happened?

Digital assessment creates more operational data, but more data does not automatically mean more visibility. When an exam event, exception, review, or appeal needs to be investigated, can the program retrieve the relevant records and understand what happened? Traceability becomes especially important when multiple systems, people, and delivery locations are involved.

3. Can controls remain consistent as delivery expands?

Certification programs increasingly operate across test centers, remote environments, and multiple markets. Growth can fragment processes if each delivery model develops its own controls, permissions, and exception handling. The challenge is to expand reach while keeping the underlying workflow understandable and consistently managed.

The role of assessment technology

Technology does not create accreditation, and no platform can determine whether a certification body meets ISO/IEC 17024. What technology can do is make good governance easier to operate and demonstrate. A well-designed assessment environment can help organizations maintain control, visibility, and evidence as their programs modernize without becoming another black box in the certification process.

That is the principle behind how ATS approaches assessment delivery.

✓ Keep AI visible — and people accountable

Where AI-assisted monitoring or scoring is used, ATS can surface indicators, anomalies, or other information for human review rather than treating automated output as the final certification decision.

The distinction is not simply whether AI is present, but whether its role in the workflow is clear. Reviewers need to understand what the system identified, retain responsibility for judgment and escalation, and be able to return to the underlying evidence when closer review is needed.

✓ Make the assessment process easier to reconstruct

When something unusual happens during an examination, the harder question often comes later: what exactly happened, who reviewed it, and what information was available at the time?

ATS supports recorded exam activity and reviewable operational records so program teams can investigate exceptions without relying solely on fragmented manual handoffs or individual recollection. As assessment processes become more digital, traceability turns operational data into usable evidence instead of simply creating more data.

✓ Maintain control as delivery becomes more complex

Certification programs rarely stay in one place. They may expand into new test centers, introduce remote delivery, enter new markets, or serve larger candidate populations.

ATS supports centralized administration, role-based permissions, identity verification, controlled access, and standardized exam workflows across delivery models. This helps program teams maintain a consistent operating model as complexity grows.

Scale is not only a question of how many candidates a system can support. It is also whether the same controls, responsibilities, and review processes remain clear as the program grows.

Keeping trust at the center

ISO/IEC 17024:2026 introduces new language and a clearer place for AI, but the underlying objective of personnel certification has not changed: stakeholders need confidence that certification decisions are based on a controlled, fair, and reliable process. As certification bodies review the new edition and their accreditation-body timelines, the question is not simply whether they have adopted new technology, but whether their assessment environment provides the visibility, human accountability, and operational control needed to use it responsibly.

ATS works with certification organizations to modernize assessment delivery through secure, traceable, and scalable workflows. If your organization is reviewing how AI, traceability, or multi-model delivery fits into its certification operations, ATS can help assess whether the underlying assessment environment is ready to support that change.

Start a conversation with us: info@atstests.com