High-risk domainAnnex III, point 3

AI in admission, assessment and examination: fair and demonstrable

AI that decides on admission, learning outcomes or exam monitoring affects students opportunities and is high-risk. We make it demonstrable that the system is fair, explainable and under human oversight.

Detail per situation

Admission and assessment of learning outcomes

AI that determines admission or assessment must be free of unfair bias and explainable. We test that and record the evidence.

Proctoring and exam monitoring

AI monitoring during exams affects privacy and equal treatment. We set up the safeguards and make compliance review-ready.

Frequently asked questions

Does an adaptive learning platform also count as high-risk?

Only if it co-decides on admission, assessment or progression with consequences. In the risk check we determine that per feature and record the rationale.

We are an edtech vendor, what do we need to demonstrate?

As a provider you carry your own obligations. We deliver the evidence pack that lets institutions buy your product without concern.

Sources and status

The legal background and sources for this topic live on Responsible AI Platform, the knowledge layer on the EU AI Act.

Read the background on Responsible AI Platform

Book a risk check for your education AI