European UnionAI Act, Article 25 (responsibilities along the AI value chain)
Regulation (EU) 2024/1689, Article 25
An AI risk obligations rule binding public and private bodies.
- How the hook was established
- express
- What makes it apply
- Market targeting, Operator establishment, Place of effect
- Obligation class
- Governance, Contract terms
- Audit expectation
- none
What it requires
- This duty takes effect on for a high-risk AI system classified under Article 6(2) and Annex III, and on for one classified under Article 6(1) and Annex I (Article 113, third paragraph, point (c), as replaced by Regulation (EU) 2026/1744), other than a system related to a product covered by the Union harmonisation legislation listed in Section B of Annex I, to which, under Article 2(2), only Article 6(1), Article 60a and Articles 102 to 112 apply, and Articles 57, 58 and 59 apply only in so far as the requirements for high-risk AI systems under this Regulation have been integrated in that Union harmonisation legislation; a system already placed on the market or put into service before its date is reached only once its design changes significantly, or by if it is intended for use by public authorities.
- If you are a distributor, importer, deployer, or other third party who puts your name or trademark on a high-risk AI system already on the market, makes a substantial modification to one that keeps it high-risk, or changes an AI system's intended purpose so it becomes high-risk, you are treated as its provider and take on the Article 16 provider obligations.
- If you are the initial provider and one of those events happens, closely cooperate with the new provider: make available technical documentation sufficient to assess compliance with Article 16, inform them of known limitations and failure modes, and give them targeted technical access, including for testing and validation, unless you clearly specified that your system was not to be changed into a high-risk one.
- If you are a product manufacturer and a high-risk AI system is a safety component placed on the market or put into service under your name or trademark, you are treated as its provider.
- If you are the provider of a high-risk AI system, or a third party supplying an AI system, AI model, tool, service, component, or process it integrates, specify by written agreement the necessary information, capabilities, technical access, and other assistance needed for the provider to comply with this Regulation, unless the third party makes the tool, service, process or component available to the public under a free and open-source licence; that exception never covers a general-purpose AI model.
These lines are LexLint's reading of the law's text.
In the frameworks
- NIST AI RMF
- GOVERN 6.1Policies and procedures are in place that address AI risks associated with third-party entities, including risks of infringement of a third-party’s intellectual property or other rights.
- NIST AI 600-1
- GAI-RISK-12Value Chain and Component Integration
- MIT mitigations Draft
- 3.3Access Management
- FINOS AIGF
- AIR-PREV-007Legal and Contractual Frameworks for AI Systems
- ETSI EN 304 223
Our reading of where this law's duties sit in each framework. It never says that running a control meets the law. How we read a framework against the law
Identifiers refer to ETSI EN 304 223 V2.1.1 (2025-12), © ETSI 2025, all rights reserved. ETSI owns the standard; no ETSI text (titles or sentences) is reproduced here. The standard is free to download from ETSI.
If you get it wrong
Criminal exposureNo
Private right of actionNo
Penalty structure
Article 99(4), point (da), inserted by Regulation (EU) 2026/1744, fines non-compliance with the obligations of providers and operators under Article 25(2) and (4) up to EUR 15,000,000 or 3% of worldwide annual turnover, whichever is higher. Article 99(6) requires the lower of the two for an SME; Article 99(6a), also inserted by that Regulation, requires the lower of the two for a small mid-cap enterprise, since both apply to paragraph 4 fines.
- Rule
- Lower of for SME
- As of
- Currency
- EUR
- Fixed cap
- 15,000,000
- Turnover percentage cap
- 3
Who enforces it
Enforcement body
National market surveillance authorities, at least one designated by each Member State under Article 70(1); the AI Office for an AI system within its exclusive competence under Article 75(1); the European Data Protection Supervisor for Union institutions, bodies, offices and agencies within the Regulation's scope, except the Court of Justice acting in its judicial capacity (Article 74(9)).
Settledness
- As of
- Open questions
- Does retraining a high-risk AI system on a deployer's own data count as a 'substantial modification' under Article 25(1), point (b), that makes the deployer the provider?
What it makes you log
Who may demand the log
Counterparty
What the log must hold
System identity
Logging duty
Paragraph 2 requires the initial provider to make available technical documentation sufficient to assess Article 16 compliance and to inform the new provider of known limitations and failure modes. It never uses the words logs, records, or audit trail, but the duty cannot be met without documenting the system's design and known limitations.
- Kind
- Implicit
- As of
- Provision
- Article 25(2)
Also on the record
EEA status
- Status
- Pending
- Source link
- https://www.efta.int/eea-lex/32024r1689
What this law does
Article 25 decides who counts as the provider of a high-risk AI system along its supply chain. A distributor, importer, deployer, or other third party is treated as the provider, and takes on the Article 16 provider obligations, if it puts its own name or trademark on a high-risk system already on the market, makes a substantial modification that keeps it high-risk, or changes an AI system's intended purpose so it becomes high-risk.
As amended by the Digital Omnibus on AI (Regulation (EU) 2026/1744, in force since ), paragraph 2 then requires the initial provider to closely cooperate with the new provider, in particular by making available technical documentation sufficient to assess Article 16 compliance, informing the new provider of known limitations and failure modes, and providing targeted technical access, including for testing and validation, unless the initial provider clearly specified that its system was not to be changed into a high-risk one.
A product manufacturer is treated as the provider where a high-risk AI system is a safety component placed on the market or put into service under the manufacturer's own name or trademark.
Paragraph 4, also amended by the Digital Omnibus to add 'AI model' to the list of what a third party may supply, requires the provider of a high-risk AI system and a third party supplying an AI system, AI model, tool, service, component, or process the system integrates to specify by written agreement the necessary information, technical access, and other assistance, unless the component is supplied under a free and open-source licence by a party other than a general-purpose AI model provider.
The Digital Omnibus inserted Article 99(4), point (da), fining non-compliance with the paragraph 2 and paragraph 4 obligations up to EUR 15,000,000 or 3% of worldwide annual turnover, whichever is higher, with the lower of the two for an SME or a small mid-cap enterprise.
Article 25 sits in Chapter III, Section 3, so like Articles 12, 19, 21 and 26 it takes effect on the schedule the Digital Omnibus wrote into Article 113: for a system classified as high-risk under Article 6(2) and Annex III, and for one classified under Article 6(1) and Annex I. Under Article 111(2), as the Digital Omnibus on AI rewrote it, the Regulation reaches the operators of a high-risk AI system placed on the market or put into service before the date Chapter III applies only if, from that date, the system's design changes significantly; in any case, the providers and deployers of a high-risk AI system intended to be used by public authorities must take the necessary steps to comply with the Regulation's requirements and obligations by .
Article 25 does not apply to a high-risk AI system related to a product covered by the Union harmonisation legislation listed in Section B of Annex I, to which, under Article 2(2), only Article 6(1), Article 60a and Articles 102 to 112 apply, and Articles 57, 58 and 59 apply only in so far as the requirements for high-risk AI systems under this Regulation have been integrated in that Union harmonisation legislation.
When LexLint raises it
When your app profile says your app makes high-risk automated decisions, distributes a software product or trains models.
Back to the example · Lint your app
Page updated .