🤖 Artificial Intelligence (AI) in Assessment
How artificial intelligence is transforming assessment design, delivery, marking, and analysis across education and professional certification.
Example in Practice
A learner uses an AI-based prompt helper during revision to practise structuring answers and checking terminology. That use may improve participation without affecting the assessed standard. If the same tool is allowed to generate phrasing during the live assessment, the team now needs to decide whether the task is still measuring the learner’s own capability or a mixed human-plus-tool performance.
FAQs
### How is identity verification different from assessment validity? Identity verification can strengthen confidence about who took the assessment, but it does not by itself prove that the assessment evidence is valid, independent, or fair. ### Can AI-based authentication reduce cheating? It can make impersonation and some forms of session misuse harder. It should still be treated as one control within a broader assessment-security design, not as a complete solution. ### What should assessment…
Suggested Citation
Test Community Network. "Generative AI in credentialing assessments." TCN AI & Assessment Wiki. Last reviewed 2026-05-02. https://www.testcommunity.network/wiki/generative-ai-in-credentialing-assessments.html
What Experts Agree On
The source set broadly converges on a cautious pattern: generative AI can help with efficiency, analysis, and some forms of assessment support, but credentialing organisations still need human control and explicit ethical handling. The strongest reading of the evidence is that AI should support the credentialing system, not quietly redefine what the credential means. There is also broad agreement that safety and ethics are not optional extras. Jisc’s primer and the GOV.UK product safety expecta…
What Is Contested
The main open question is how far generative AI can be used inside credentialing without changing the award’s meaning. Vendor and practitioner material often frames AI as a route to better efficiency and accuracy, but that does not settle where the line should sit between support and automation. The unresolved issue is whether the system is simply faster, or actually more defensible. A second tension is between student-facing ethics guidance and operational credentialing practice. Coursera’s et…
Definition
This topic covers the use of generative AI in credentialing assessments: certification, licensure, professional testing, and other awards where a result carries formal recognition or regulatory weight. In practice, AI may support item development, scoring support, analytics, content review, candidate guidance, or moderation. The assessment issue is whether those uses preserve the construct being measured and the defensibility of the resulting credential.
Key Concepts
- **Identity assurance**: confidence that the candidate is who they claim to be. - **Authentication**: the mechanism used to confirm or re-confirm identity during a session. - **Biometrics**: identity checks using facial, voice, behavioural, or similar signals. - **Continuous authentication**: repeated checks during the assessment rather than a single login event. - **Assessment validity**: whether the evidence collected is actually the evidence needed for the decision. A system can be strong o…
Vendor Landscape
Vendor and platform material tends to frame AI as supporting accessibility, personalisation, and consistent learner experience. That is a useful market signal, but it does not settle whether a tool is fair in a particular assessment context. Claims about inclusion should therefore be treated as prompts for evaluation rather than proof of fair outcomes.
Example in Practice
A professional certification body wants to use generative AI to help draft item variants and summarise item performance. That can be defensible if subject experts keep approval authority, bias checks are documented, and no AI system is allowed to award the credential itself. The same body would need much stronger evidence before letting AI influence pass/fail decisions directly.
Last Reviewed By
Tim Burnett (Admin)