Glossary
Deployer (EU AI Act) — Definition for Lawyers
A deployer under the EU AI Act is anyone using an AI system under its own authority for professional purposes. Law firms are deployers; here is what follows, and when.
On this page
A deployer, under Article 3(4) of the EU AI Act, is any natural or legal person, public authority, agency or other body that uses an AI system under its own authority, except in a personal, non-professional activity. A law firm that runs ChatGPT, Copilot, Harvey or an in-house tool is a deployer of each. The label decides which obligations reach the firm and from which date.
Provider, deployer, distributor: three roles, one firm
The AI Act allocates duties by role, not by industry. The heavy obligations sit with the provider, the party that develops an AI system or has it developed and places it on the market or puts it into service under its own name (Article 3(3)). Importers and distributors cover third-country and supply-chain cases. The deployer is the user. One firm can hold different roles for different tools on the same day.
| Role | Statutory test | Law-firm example | Practical weight |
|---|---|---|---|
| Provider | Develops the system, or has it developed, and markets it or puts it into service under its own name | A firm that builds its own tool for internal use, such as Kirkland's planned $500m platform (reported May 2026) | Full obligations if the system is high-risk; documentation and copyright duties if it is a general-purpose model |
| Deployer | Uses the system under its own authority in a professional context | Every firm using ChatGPT, Copilot, Harvey, Legora or CoCounsel | Article 4 literacy now; Article 26 oversight and logging for high-risk systems from December 2027 |
| Distributor | Makes a system available on the EU market without being provider or importer | A firm reselling a vendor's agents to clients under the vendor's name, the model A&O Shearman adopted with four Harvey agents in April 2025 | Verify marking and documentation for high-risk systems; otherwise light |
| Importer | Places on the EU market a system bearing a non-EU provider's name | Rare for law firms | Check conformity before import |
When a deployer becomes the provider
Article 25 converts a deployer into the provider of a high-risk system in three situations: it puts its own name or trademark on a high-risk system already on the market, it makes a substantial modification to one, or it changes the intended purpose of a system so that it becomes high-risk. For firms this is the live question in build-versus-buy decisions. Latham & Watkins is fine-tuning Nvidia's open-weight Nemotron 3 models on its own H200 servers (September 2026); whether that makes the firm a provider of anything depends on what the resulting system is used for, not on the engineering. A research assistant for the firm's own lawyers is not an Annex III use. An HR screening tool is.
Deployer duties by tier and date after the Omnibus
The Digital Omnibus on AI, Regulation (EU) 2026/1744, entered into force on 27 July 2026 and moved the high-risk dates without touching the deployer definition. The table gives the position as of September 2026.
| Duty | Basis | Applies to a law-firm deployer? | Date |
|---|---|---|---|
| Avoid prohibited practices, including two new limbs on non-consensual intimate imagery | Art. 5 | Yes, for any system | Since 2 Feb 2025; new limbs by 2 Dec 2026 |
| Support staff AI literacy; keep a training record | Art. 4, reworded by the Omnibus into an obligation of effort | Yes, for any organisation whose staff use tools such as ChatGPT | Since 2 Feb 2025; supervised by national authorities since 2 Aug 2026 |
| Transparency: chatbots identify as AI; synthetic content is labelled | Art. 50 | Yes where the firm runs a client-facing chatbot or publishes AI-generated content | 2 Aug 2026; watermarking grace to 2 Dec 2026 for existing systems |
| Use per instructions, assign human oversight, ensure relevant input data, monitor, keep logs, inform workers | Art. 26 | Only for high-risk systems: HR screening and worker management; possibly research tools "assisting judicial authorities", which is unsettled | 2 Dec 2027 (Annex III); 2 Aug 2028 (Annex I) |
| Fundamental rights impact assessment | Art. 27 | Public bodies and public-service providers; not a private firm advising clients | 2 Dec 2027 |
Two points deserve emphasis. Article 4 is the duty with teeth today: national market surveillance authorities have supervised it since 2 August 2026, and the Commission's FAQ (updated 27 July 2026) confirms it covers any organisation whose staff use tools like ChatGPT; there is no dedicated EU fine tier, but national penalties apply. And the deferral is a deferral, not a cancellation: Gibson Dunn's summary of the Omnibus is the cleanest account of what moved and what did not. In Austria the KI-Servicestelle at RTR is the national information hub.
What to do this week
Build a one-page register: every AI system the firm uses, the vendor, the firm's role for it, whether any use touches Annex III (recruitment and worker management are the usual catch), and the date of the last literacy session. Anthropic's ai-governance-legal plugin produces exactly this per-system inventory, and the AI literacy guide covers the training record. Module 6 of the AI Lab for Lawyers course walks through the role analysis tool by tool. For the wider picture read the AI Act for law firms and what the Omnibus changed; for the model-level obligations that sit with your vendor rather than with you, see general-purpose AI model; employment exposure is covered in AI for employment lawyers.
Frequently asked questions
Is a law firm a deployer or a provider under the EU AI Act?
Almost always a deployer. Using ChatGPT, Copilot, Harvey or Legora under the firm's authority makes the firm a deployer of each system, and the vendor is the provider. The firm becomes a provider only if it develops its own system, puts its own name on a high-risk system, substantially modifies one, or changes its intended purpose, as Article 25 sets out.
What does a deployer have to do right now, in September 2026?
Support AI literacy among staff under Article 4, avoid the Article 5 prohibited practices, and observe Article 50 transparency where the firm runs chatbots or publishes AI-generated content. High-risk deployer duties under Article 26, such as human oversight and log retention, do not apply until 2 December 2027 for Annex III systems following the Digital Omnibus.
Does a deployer need a fundamental rights impact assessment?
Only in narrow cases. Article 27 requires a fundamental rights impact assessment from public bodies, private entities providing public services, and deployers of certain credit and insurance systems, before using a high-risk Annex III system. A private law firm advising clients does not fall within it, although a firm using AI to screen job applicants should still document its oversight arrangements.
Put this into practice
AI Lab for Lawyers: a live, hands-on course for practising lawyers
4 live sessions (8 hours), recordings included, taught by Dr. Niklas Schmidt of Wolf Theiss. Rated 4.7/5 from 381 ratings; alumni from Norton Rose Fulbright, Paul Hastings, Pillsbury and Taylor Wessing.
Related reading
Regulation & Ethics
The EU AI Act for Lawyers and Law Firms: What Applies in 2026
The EU AI Act applies to law firms as deployers. Article 4 literacy and Article 50 transparency are live; high-risk duties are deferred to December 2027.
9 min read
Regulation & Ethics
The Digital Omnibus on AI: What Changed in the AI Act and What Did Not
The Digital Omnibus AI Act amendment, Regulation (EU) 2026/1744, deferred high-risk duties to December 2027 but left transparency, literacy and the prohibitions in force.
6 min read
Regulation & Ethics
EU AI Act Article 4: The AI Literacy Duty for Law Firms, Explained
Article 4 of the AI Act requires every firm whose staff use AI to support their AI literacy. In force since 2 February 2025, supervised nationally since August 2026.
7 min read
Questions
Does the EU AI Act Apply to Law Firms? A Direct Answer
Yes. An EU law firm whose staff use ChatGPT, Claude or Copilot is a deployer under the EU AI Act. Article 4 literacy applies now, Article 50 since August 2026.
6 min read
Glossary
General-Purpose AI Model (GPAI) — Definition for Lawyers
A general-purpose AI model under the EU AI Act is a model with significant generality, such as GPT-6 or Claude Opus 5. Definition, tiers, signatories, enforcement.
4 min read
Practice Areas
AI for Employment Lawyers: Advice, Investigations and the AI Act's HR Rules
AI for employment lawyers: advising on AI in hiring and monitoring after the AI Act's Annex III deferral to December 2027, plus investigations and policies.
11 min read