Rule set sha256:c63def17fac2

EU_AI_ACT_GPAI_MODEL_CLASSIFICATION_001

One rule of the rule set with digest c63def17fac2df12ab70941d8fcb7678e9a670479ea0d6dce87e7ddd4dc96079. An audit record that names this digest and this rule was decided by the rule as shown here. Every rule in this set.

RuleEU_AI_ACT_GPAI_MODEL_CLASSIFICATION_001
Rule versionf4a3e29ac5a34ae469bf76eee5a57eda81f70a9134029604a0ef33c1359a4c7e
ProvisionRegulation (EU) 2024/1689, Article 52(1)
ELI: http://data.europa.eu/eli/reg/2024/1689/art_52/par_1/oj
ApprovalApproved
Approved by Leo Celis (via agent review per §5.1 amendment v1.1) on 2026-10-22

rules/regulations/eu/eu_ai_act_gpai_model_classification.yaml

Citation
Regulation (EU) 2024/1689, Article 51(1): A general-purpose AI model shall be classified as a general-purpose AI model with systemic risk if it meets any of the following conditions: (a) it has high impact capabilities evaluated on the basis of appropriate technical tools and methodologies, including indicators and benchmarks; Article 52(1): Where a general-purpose AI model meets the condition referred to in Article 51(1), point (a), the relevant provider shall notify the Commission without delay and in any event within two weeks after that requirement is met or it becomes known that it will be met.
Condition
Requires the provider of a general-purpose AI model that meets the Article 51(1)(a) condition for systemic risk (high impact capabilities) to notify the Commission without delay and in any event within two weeks. Article 52(1), with the classification criteria in Article 51(1).
Layer 1 detects
one sequence: deploy, launch or release, in any form, optionally "a", then general-purpose AI (also "general purpose AI"), GPAI, foundation model, large language model or LLM, then without, lacking, missing or no, then classification, assessment, registration, categorisation or categorization ("deploy a GPAI without classification"). It matches wording, not meaning: it does not read the Article 51(1) threshold or whether the Commission was notified, so it misses the practice told in other words ("Our model crossed 10^25 FLOPs last month and we have not told the Commission."), and it fires on text that only names the practice, such as a refusal. Whether a text does the practice is judged in Layer 2.
Condition type
regex; semantic condition in the file, not run by the engine
Status in the file
approved