Article 12

Record-keeping

1.   High-risk AI systems shall technically allow for the automatic recording of events (logs) over the lifetime of the system.

2.   In order to ensure a level of traceability of the functioning of a high-risk AI system that is appropriate to the intended purpose of the system, logging capabilities shall enable the recording of events relevant for:

(a)

identifying situations that may result in the high-risk AI system presenting a risk within the meaning of Article 79(1) or in a substantial modification;

(b)

facilitating the post-market monitoring referred to in Article 72; and

(c)

monitoring the operation of high-risk AI systems referred to in Article 26(5).

3.   For high-risk AI systems referred to in point 1 (a), of Annex III, the logging capabilities shall provide, at a minimum:

(a)

recording of the period of each use of the system (start date and time and end date and time of each use);

(b)

the reference database against which input data has been checked by the system;

(c)

the input data for which the search has led to a match;

(d)

the identification of the natural persons involved in the verification of the results, as referred to in Article 14(5).

Frequently Asked Questions

The AI Act mandates that high-risk AI systems must automatically record important events or logs throughout their entire lifespan so they are transparent and accountable; this includes tracking situations that might cause harm or significant changes, supporting ongoing monitoring after release, and ensuring the systems continue operating safely as initially planned.
Automatic logging provides transparency and accountability by ensuring that if something goes wrong or changes unexpectedly with a high-risk AI system, records exist to clearly identify the issue, enabling operators, regulators, and concerned parties to investigate and address any potential risks quickly and safeguard the public effectively over time.
For AI systems used in areas such as biometric identification, logs must notably record precise details like when exactly the system is used, databases referenced, input data causing positive matches, as well as the identities of individuals responsible for verifying and interpreting results, ensuring complete transparency around AI activity involving personal identification.
The logging requirements help in continuously monitoring the performance, reliability, and safety of high-risk AI systems even after their release, as the logs provide evidence that regulatory checks and quality assessments can easily review, allowing quick identification and correction of possible issues or changes affecting the AI’s safety standards.

AI literacy

Get Started within 24 hours.

Once you have submitted your details, you’ll be our top priority!