ANNEX I
Global AI Risk Register
The Global AI Risk Register records risk classification determinations made under Annex I: AI systems and classes of systems assigned to a risk tier following the criteria of Annex I, Part 1, including assignment to the High-Risk areas enumerated in Part 2. The classification framework itself is published with the risk classification annex.
Classification is not a judgment on the quality of a system. It determines which obligations of the framework apply to it: the higher the tier, the more demanding the certification, audit and monitoring requirements that attach.
REVIEW
Reclassification
Each entry records the system or class, the provider or operator, the area engaged, the tier assigned, the date of the determination and the next scheduled review. Determinations are reviewed periodically and on material change: a new capability, a change of scale, or deployment in a new context can move a system between tiers.
A superseded determination remains visible with its status recorded, so the classification history of a system can be traced. Classification determinations are appealable under Article 43 of the Statute.
Consult the Register
Entries are published once the underlying decision has become final or is immediately effective. Requests for the correction of an entry may be submitted through the contact page; a request does not suspend the entry.
