As of 2026-10-07, 13 official documents govern AI use by lawyers, judges or litigants in the European Union. 6 require disclosure or certification of AI use, 2 impose an explicit verification duty, and 3 prohibit some use outright. The incident tracker records 1 incident in this jurisdiction.
The EU AI Act sets risk-based obligations for providers and deployers of AI systems, including law firms and legal tech vendors operating in the European Union. Prohibited practices and AI literacy duties applied from 2 February 2025, GPAI rules from 2 August 2025 and Article 50 transparency from 2 August 2026, with high-risk obligations now deferred to 2027 and 2028.
Regulation (EU) 2026/1744, adopted 8 July 2026 and in force from 27 July 2026, amends the EU AI Act, the EASA Regulation and the Machinery Regulation. It defers Annex III high-risk obligations to 2 December 2027 and Annex I obligations to 2 August 2028. Article 50 transparency and GPAI duties were not deferred and applied from 2 August 2026.
European Commission guidelines, approved 20 July 2026 as C(2026) 5054 final, on the EU AI Act Article 50 transparency duties that applied from 2 August 2026. Providers mark synthetic output in a machine-readable, detectable way; deployers disclose deepfakes and AI-generated text on matters of public interest, unless the text has had substantive human review.
A voluntary EU code, finalised 10 June 2026 and confirmed adequate by the Commission in July 2026, that shows providers and deployers how to meet the AI Act's Article 50 marking and labelling duties for AI-generated text, images, audio and video. Around 190 signatories had joined by late July 2026, with watermark-detection interoperability due by 2 February 2027.
Draft guidelines published by the European Commission on 19 May 2026 on classifying AI systems as high-risk under Article 6 of the EU AI Act, including administration-of-justice uses listed in Annex III. A targeted consultation closed in summer 2026 and final guidelines are expected by the end of 2026, ahead of the deferred 2 December 2027 application date.
Companion to the CCBE generative-AI guide and cloud-computing guidelines. It explains deployment options - on-premises, self-hosted colocation, bring-your-own-model and fully managed SaaS - and their confidentiality and data-residency implications, model capability, size, speed and context-length considerations, which models can run locally, and a technical glossary grounding the duty to understand the tools used.
A charter adopted by the Court of Justice of the European Union on 26 January 2026 governing AI tools developed or deployed inside the institution, including tools that assist the Court of Justice and the General Court in their judicial functions. It names automation bias as a specific risk to case-law development and judicial quality.
Guide published 2 October 2025 by the CCBE, which represents European bars and law societies, on how lawyers should use generative AI. It is permissive with cautions: lawyers must understand tool limitations, verify all output, protect professional secrecy and client confidentiality, and consider the EU AI Act's literacy and transparency duties.
European Commission guidelines published 18 July 2025 clarifying which organisations are providers of general-purpose AI models under the EU AI Act and what documentation, copyright and transparency duties attach from 2 August 2025. The guidelines are interpretive and relevant to legal tech vendors that fine-tune or distribute foundation models in the EU.
A voluntary code drafted under the EU AI Office and published 10 July 2025 that gives providers of general-purpose AI models a route to demonstrate compliance with the AI Act's GPAI obligations. Its chapters cover transparency, copyright and safety and security. Commission enforcement powers over GPAI providers became effective on 2 August 2026.
European Commission questions and answers, published May 2025, on the Article 4 AI literacy duty of the EU AI Act that applied from 2 February 2025. Deployers, including law firms that use AI systems in the EU, must take measures to support AI literacy for staff and others dealing with those systems.
Interpretive guidelines from the European Commission, published 4 February 2025, explaining the prohibited AI practices in Article 5 of the EU AI Act, which applied from 2 February 2025. The guidelines are non-binding but set out how the Commission reads the bans on manipulative techniques, social scoring, untargeted facial scraping and emotion recognition in workplaces.
An EDPB opinion adopted on 17 December 2024 answering when an AI model can be treated as anonymous, how controllers justify legitimate interest at the development and deployment stages, and what follows if training data was processed unlawfully. It matters to law firms and legal-tech vendors deploying AI on personal data in the EU.
Descriptive reference only: this page records what each rule requires. Missing a document? Tell us.
Reuse this register
Show the latest rules in the European Union on your own site with a widget, run the weekly block in your newsletter, or pull the JSON. Everything links back to the versioned record; data CC BY 4.0.
European Union