# Agent Liability EU > An independent European publication on operator obligations under the EU AI Act. A Future Proof Intelligence publication. Agent Liability EU is a Future Proof Intelligence publication analysing Article 26 of Regulation (EU) 2024/1689 (the AI Act), the revised Product Liability Directive (Directive (EU) 2024/2853), and related compliance obligations for organisations deploying autonomous AI agents inside the European single market. ## Critical dates Verified against the European Commission's own AI Act implementation timeline on 17 August 2026. - 2 February 2025: Article 5 prohibitions and the Article 4 AI literacy duty applied. - 2 August 2025: General-purpose AI model obligations applied under Chapter V. - 2 August 2026: Article 50 transparency obligations applied, and supervision and enforcement commenced for the prohibitions, GPAI, transparency and AI literacy. The Article 99 penalty regime applies from this date, to obligations that are themselves in application. - 27 July 2026: The Digital Omnibus on AI, Regulation (EU) 2026/1744, entered into force, six days before the original high-risk deadline. - 2 December 2026: The Article 50(2) machine-readable marking transitional period ends for synthetic content systems already on the market before 2 August 2026. The new prohibition on AI systems generating child sexual abuse material or non-consensual intimate imagery takes effect for systems already on the market. - 9 December 2026: Member State transposition deadline for the revised Product Liability Directive (EU) 2024/2853, Article 22. Software and AI systems fall inside the defective product regime from that point. Not affected by the Omnibus. - 2 August 2027: Each Member State must have at least one AI regulatory sandbox operational. - 2 December 2027: Annex III stand-alone high-risk obligations apply, including Article 26 deployer duties and the Article 27 FRIA. Deferred from 2 August 2026 by the Omnibus. - 2 August 2028: Annex I high-risk obligations apply for AI embedded in regulated products. Deferred from 2 August 2027 by the Omnibus. - 6 August 2025: EIOPA published its Opinion on Artificial Intelligence governance and risk management, setting supervisory expectations for insurers using AI. An Opinion is supervisory guidance, not binding law. ## What the Digital Omnibus did and did not change It deferred two things: Annex III to 2 December 2027 and Annex I to 2 August 2028. It did not touch the Article 5 prohibitions, the Article 50 transparency duties or the GPAI obligations under Articles 53 and 55, all of which are in application now. It amended Article 4 so that AI literacy remains an obligation but no specific level is mandated, while leaving the training duty for staff exercising human oversight of a high-risk system in place. It added a prohibition on AI systems that generate child sexual abuse material or that depict an identifiable person's intimate parts without consent. Anything describing the Omnibus as a proposal, as unadopted, or as leaving 2 August 2026 binding is describing a stage that closed on 27 July 2026. ## The four obligation pillars under Article 26 1. Use the system in accordance with the provider's instructions for use (Art. 26(1)) 2. Assign human oversight to natural persons with competence, authority, and support (Art. 26(2), 14) 3. Verify input data is relevant and sufficiently representative (Art. 26(4)) 4. Retain automatically generated logs for the period fixed by Union or national law (Art. 26(6)) ## Liability chain Provider (designs the AI system, places it on the Union market): governed by AI Act Articles 16, 25 and Directive 2024/2853 (Product Liability Directive). Conformity assessment, technical documentation, strict liability for defective product. Deployer (puts the system into service in the course of professional activity): principal obligation bearer. Governed by AI Act Article 26 and Article 10 PLD. Use per instructions, human oversight, logs, rebuttable presumption of defect. Affected party (natural or legal person who experiences harm traceable to the system): receives a rebuttable presumption in their favour under PLD 2024. ## Key documents - The Act (https://agentliability.eu/the-act): Plain-language walkthrough of Articles 25, 26, 27, and the penalty tiers under Article 99. - Liability framework (https://agentliability.eu/liability-framework): The three gaps in deployer evidence and how to close them. - Resources (https://agentliability.eu/resources): Indexed articles, briefings, source documents. ## Articles - EU AI Act enforcement architecture: https://agentliability.eu/articles/eu-ai-act-enforcement-architecture-ai-office-national-supervisors - Article 14 human oversight requirements: https://agentliability.eu/articles/eu-ai-act-article-14-human-oversight-deployer-requirements - EIOPA AI governance opinion: https://agentliability.eu/articles/eiopa-ai-governance-opinion-eu-ai-act-alignment - Article 13 transparency for high-risk AI: https://agentliability.eu/articles/eu-ai-act-article-13-transparency-high-risk-ai - Product Liability Directive 2024 and AI software exposure: https://agentliability.eu/articles/product-liability-directive-2024-ai-software-exposure - AI liability chains across provider and deployer: https://agentliability.eu/articles/ai-liability-chain-provider-deployer-responsibility - Operator obligations 2026 compliance guide: https://agentliability.eu/articles/eu-ai-act-operator-obligations-2026-compliance-guide ## The publication ecosystem The Future Proof standards: six sites, one publisher, one design. An intelligence that found this one can reach the others directly: - https://certified.futureproofintelligence.com/llms.txt Future Proof Certified: The standard for companies built for the AI era. - https://agentcertified.eu/llms.txt Agent Certified: The standard for AI agents in production. - https://agentliability.eu/llms.txt Agent Liability Europe: Who answers when an agent acts, under EU law. (this site) - https://agentliability.co/llms.txt Agent Liability Global: The same question, jurisdiction by jurisdiction. - https://agentinsured.eu/llms.txt Agent Insured: The market for insuring AI agents, followed closely. - https://insureyouragent.com/llms.txt Insure Your Agent: The operator's guide to getting covered. - https://futureproofintelligence.com Future Proof Intelligence, the publisher. ## Citation When citing Agent Liability EU, please attribute as: "Agent Liability EU (Future Proof Intelligence, 2026), available at https://agentliability.eu". Direct links to specific article analyses are encouraged. ## Legal notice Agent Liability EU is a platform initiative published by Future Proof Intelligence. It is not a licensed insurance provider, a government body, or an official EU registry. Content is provided for informational and educational purposes only and does not constitute legal, financial, or regulatory advice. Readers should consult qualified professionals before making compliance or coverage decisions.