Article 101: Fines for Providers of General-Purpose AI Models
Article 101 creates a dedicated fine regime for providers of general-purpose AI (GPAI) models — separate from the general penalty framework in Article 99. The AI Office (not national authorities) can impose fines for Chapter V infringements. Maximum fines are EUR 15 million or 3% of worldwide annual turnover for non-compliance with GPAI obligations, and EUR 7.5 million or 1% for supplying incorrect information. The Commission adopts procedural rules for investigations and fine decisions. Always verify on EUR-Lex.
Who does this apply to?
- -Providers of GPAI models subject to Chapter V obligations (Articles 53–55)
- -Providers of GPAI models with systemic risk under Article 51
- -The AI Office as the enforcing authority for GPAI fines
- -Legal and compliance teams modelling GPAI-specific fine exposure
Scenarios
A GPAI model provider fails to prepare Annex XI technical documentation as required by Article 53(1)(a).
A systemic-risk provider supplies misleading information to the AI Office about its model's training compute when responding to a formal request.
A GPAI provider also deploys a high-risk system in the EU. Authorities investigate both the GPAI and high-risk aspects.
The two GPAI fine tiers (plain terms)
Article 101 creates two tiers of fines specifically for GPAI provider infringements:
Tier 1 — Non-compliance with GPAI obligations: - Up to EUR 15 million or 3% of total worldwide annual turnover, whichever is higher - Covers: Article 53 (baseline GPAI duties), Article 55 (systemic risk duties), Annex XI documentation, copyright/TDM policy, downstream information obligations
Tier 2 — Incorrect, incomplete, or misleading information: - Up to EUR 7.5 million or 1% of total worldwide annual turnover, whichever is higher - Applies to information supplied to the AI Office in response to requests
These figures are ceilings — the AI Office applies criteria when setting actual amounts.
Who enforces — the AI Office
Unlike Article 99 (where national authorities enforce), Article 101 grants enforcement power to the AI Office within the Commission. This means:
- The AI Office investigates, issues requests for information, and imposes fines
- Procedural rules are adopted by the Commission (implementing acts)
- The Court of Justice of the EU has unlimited jurisdiction to review fine decisions
- National authorities remain responsible for Article 99 enforcement on high-risk/transparency obligations
Relationship to Article 99
Article 101 is the GPAI-specific counterpart to Article 99:
| | Article 99 | Article 101 | |---|---|---| | Scope | All AI Act obligations except GPAI | GPAI model obligations (Chapter V) | | Enforcer | National authorities | AI Office | | Highest tier | EUR 35M / 7% (prohibited practices) | EUR 15M / 3% | | Application date | Most from Aug 2025 | Aug 2026 |
Where a provider is both a GPAI provider and a high-risk system provider, both Article 99 and 101 may apply to different aspects of the same entity.
Recitals (preamble) on EUR-Lex
The recitals in the same consolidated AI Act on EUR-Lex contextualise the centralised enforcement model, AI Office powers, and the rationale for delayed application (obligations active since Aug 2025, fines from Aug 2026 to allow transition). Use the official preamble on EUR-Lex.
Compliance checklist
- Model GPAI-specific fine exposure separately from Article 99 high-risk exposure.
- Map each Chapter V obligation (Articles 53, 55) to its fine tier under Article 101.
- Track AI Office procedural rules for investigations and formal requests.
- Ensure accuracy and completeness of all information supplied to the AI Office.
- Coordinate GPAI fine risk with Article 99 risk for dual-obligation entities.
- Brief management on the Aug 2026 application date — obligations are active since Aug 2025 but fines only from Aug 2026.
Model your GPAI fine exposure—free assessment.
Start Free AssessmentRelated Articles
Article 53: Obligations for Providers of General-Purpose AI Models
Article 55: Obligations for Providers of GPAI Models with Systemic Risk
Article 51: Classification of GPAI Models with Systemic Risk
Article 52: Procedure for Systemic Risk Classification of GPAI Models
Article 56: Codes of Practice for GPAI Models
Article 99: Penalties for AI Act Infringements
Article 64: European Artificial Intelligence Board
Article 113: Entry into Force and Application Dates
Annex XI: Technical Documentation for Providers of General-Purpose AI Models
Related annexes
- Annex XI — GPAI technical documentation (breach triggers Article 101)
Frequently asked questions
Why is Article 101 delayed to August 2026 when GPAI obligations started August 2025?
The one-year gap is a transitional grace period. Providers must comply with Chapter V from August 2025, but the AI Office cannot impose fines until August 2026, allowing time to establish enforcement procedures.
Can national authorities fine GPAI providers?
For GPAI-specific (Chapter V) obligations, enforcement is centralised at the AI Office under Article 101. National authorities enforce Article 99 for non-GPAI obligations that may also apply to the same entity.
Are Article 101 fines on top of Article 99 fines?
They can apply concurrently to different obligations. A GPAI provider who also deploys high-risk systems faces AI Office enforcement for Chapter V and national enforcement for Chapter III. Proportionality and ne bis in idem principles apply.
Do SME caps apply to Article 101?
Article 101 sets maximum amounts proportionate to turnover. SME-specific caps are primarily in Article 99(6) — check whether Article 101 references equivalent provisions on EUR-Lex.