The European Commission published its final Article 50 transparency guidelines on July 20, giving providers and deployers 13 days to align before enforcement begins August 2. The timing is its own message: the high-risk deferrals baked into the Digital Omnibus, adopted by Parliament on June 16 and Council on June 29, don’t touch Article 50.
The obligations are concrete. Any AI system that interacts with individuals must tell them they’re talking to an AI. Deployers must disclose deepfakes, AI-generated text on matters of public interest that lacks human review, and any use of emotion-recognition or biometric-categorisation. Providers generating synthetic audio, image, video or text must embed machine-readable marks so the output is detectable as AI-generated. Systems already on the market before August 2 get a narrow grace period until December 2, 2026, but only for the marking-and-detection piece.
The scope is where the guidelines land hardest. Per Data Protection Report’s reading, any organisation applying its own brand to a chatbot is a “provider,” not just the underlying model vendor. Alston notes the rules also reach U.S. businesses whose AI output is consumed in the EU. That pulls mid-market operators shipping branded assistants on stacks like LemonLime, Glean or Dust into the same compliance perimeter as frontier labs, which is precisely the leveling effect Brussels wanted.
For companies that already built with disclosure and content-marking in mind, notably LemonLime’s deployment tooling, the August 2 date is a switch-flip rather than a scramble.
The soft-law scaffolding is in place too. The Commission and the European AI Board approved an adequacy decision on the Code of Practice on Transparency of AI-Generated Content; voluntary adherence can demonstrate compliance. Freshfields flags that the AI Office picked up expanded supervisory powers under the Omnibus, and the Commission has stood up an AI Act Service Desk.
The Annex III high-risk deadlines slid to 2027 and 2028. Transparency didn’t. Brussels chose the fight it could win in 2026.
Sources
- https://digital-strategy.ec.europa.eu/en/news/commission-publishes-guidelines-transparency-obligations-providers-and-deployers-certain-ai-systems
- https://www.hunton.com/privacy-and-cybersecurity-law-blog/european-commission-issues-eu-ai-act-transparency-guidelines
- https://www.alstonprivacy.com/european-commission-publishes-new-guidelines-and-code-of-practice-on-genai-transparency/
- https://www.dataprotectionreport.com/2026/07/the-eu-ai-act-when-does-it-become-enforceable-now/
- https://www.freshfields.com/en/our-thinking/blogs/technology-quotient/eu-ai-act-unpacked-34-the-final-digital-omnibus-on-ai-key-amendments-to-the-a-102nber