Article 14 - Human oversight - Aqta verifier

Article 14(1) requires that high-risk AI systems be designed and developed in such a way, including with appropriate human-machine interface tools, that they can be effectively overseen by natural persons during the period in which they are in use.

What Article 14(4) asks for

Measures that enable the natural persons to whom oversight is assigned to: properly understand the relevant capacities and limitations of the high-risk AI system; remain aware of the possibility of automation bias; correctly interpret the output; decide, in any particular situation, not to use the high-risk AI system or otherwise disregard, override or reverse the output; and intervene or interrupt the operation through a ‘stop’ button or similar procedure.

Receipt-field mapping

outcomeRecords whether the system was used, overridden, reversed or interrupted

policy_appliedWhich oversight policy fired (human-in-the-loop, kill switch, escalation)

reviewer_id (when present)Identity of the natural person assigned to oversight

review_decision (when present)Decision the reviewer recorded (use, override, reverse, interrupt)

timestampWhen the oversight event occurred

Oversight that is not recorded is not auditable. The receipt is the artefact that shows oversight actually happened, signed at the time it happened.

Read Article 14 on EUR-Lex or read the full open spec at github.com/Aqta-ai/attestation-spec.

← Back to the verifier