EU AI Act enforcement begins

The European Commission began enforcing parts of the EU AI Act on August 2, 2026, when Article 50 transparency obligations started applying to certain AI systems. The rules require some providers and deployers to tell people when they are interacting with AI and when content has been generated or manipulated by AI, according to the Commission. The Commission said its AI Office, working with national authorities, would start enforcement from that date. ### Which AI Act rules actually started on August 2? The AI Act entered into force on August 1, 2024, but the Commission says its main body became applicable on August 2, 2026, with some exceptions. The Commission’s policy page says prohibited AI practices and AI literacy duties had already started applying on February 2, 2025. Article 50 is the immediate focal point this month. (digital-strategy.ec.europa.eu) The Commission says those transparency rules apply to certain AI systems, including systems that interact directly with people and systems that generate or alter content, so users can recognize when AI is involved. ### What do deployers have to do now? The Commission said deployers of covered systems must ensure users are informed when they are interacting with AI, where Article 50 requires it. (digital-strategy.ec.europa.eu) The Commission’s guidelines were published to help providers, deployers and competent authorities apply those obligations in practice. DLA Piper said the change matters for employers because deployer obligations now apply from August 2, 2026, and can reach workplace uses of AI. (digital-strategy.ec.europa.eu) In DLA Piper’s account, employers using automated tools in hiring or employment decisions need to review practical steps such as transparency measures, internal documentation and compliance processes. (digital-strategy.ec.europa.eu) ### Why are employers being singled out in this story? DLA Piper said employment-related AI can fall within the AI Act’s higher-risk categories, especially where tools are used in recruitment, selection or management of work-related relationships. Its July guidance on draft high-risk guidelines said AI tools may be high-risk when they materially influence employment decisions, even if a human makes the final call. (knowledge.dlapiper.com) That means the issue is not limited to AI vendors. Companies that buy and use third-party systems can still be “deployers” under the Act and may need to show how notices are given, how the tool is used, and what records are kept, according to DLA Piper’s August 5 note. ### What does this mean for companies using AI in hiring? (knowledge.dlapiper.com) Employers using AI in recruitment may need to check whether applicants and workers are being told when AI is used in interactions or content-related functions covered by Article 50. DLA Piper said companies should also look at policies, documentation and vendor arrangements as the rules become enforceable. (knowledge.dlapiper.com) The Commission said it has also issued a voluntary Code of Practice on transparency of AI-generated content alongside the guidelines. That code is intended as a structured compliance pathway for organizations preparing to meet the Article 50 rules. ### Where is the official compliance detail? The European Commission published guidelines on transparency obligations on July 27, 2026, and says they are meant to help providers and deployers understand scope, exemptions and implementation. (knowledge.dlapiper.com) The Commission has also published a fact page and a FAQ page on Article 50. (digital-strategy.ec.europa.eu) On July 31, 2026, the Commission said its AI Office and national authorities would begin enforcement from August 2. For employers and other deployers, the next step is in those Commission guidelines and FAQs, which set out the operational detail national authorities and companies will use as the rules begin to bite. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2)