EU rules push AI transparency

Europe’s artificial intelligence rulebook is moving from broad principles to product-level disclosure duties that start applying across the bloc on August 2, 2026. (digital-strategy.ec.europa.eu) The European Union’s AI Act entered into force on August 1, 2024, with a staggered timetable: bans on certain practices and AI literacy duties started on February 2, 2025, while Article 50 transparency obligations take effect on August 2, 2026. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2) Article 50 covers four buckets of disclosure: systems that interact with people must say they are AI unless that is obvious; AI-generated or AI-manipulated content must be marked in machine-readable form; users exposed to emotion recognition or biometric categorisation must be informed; and deepfakes or AI-made public-interest text must disclose their artificial origin in defined cases. (digital-strategy.ec.europa.eu) The European Commission says those notices must be “clear and accessible,” and it is consulting on guidelines and a code of practice for transparent generative AI systems. The same Commission guidance points to tools such as watermarks, metadata, cryptographic provenance, logs and fingerprints to help mark and detect synthetic content. (digital-strategy.ec.europa.eu) That push lands alongside the Digital Services Act, which already governs online platforms used by Europeans, including online travel and accommodation platforms. The law requires explanations for content removals, labels on ads, information on who placed them and why they are shown, and non-personalised feed options on very large platforms with more than 45 million monthly users in the European Union. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2) For companies shipping consumer-facing AI in Europe, the overlap means compliance is no longer just a legal memo. It now reaches chat windows, booking flows, ad systems, content labels, audit logs and staff training programs that show workers when a model is being used and what it can and cannot do. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2) (iapp.org) Hotels are one example because they already use AI in revenue management, demand forecasting, customer service and targeted marketing. Hotel News Resource reported on April 22, 2026, that stricter European rules could force operators to document how algorithms affect pricing and guest interactions, including when systems come from outside vendors. (hotelnewsresource.com) The AI Act itself is built on a risk ladder, with separate rules for prohibited uses, high-risk systems and transparency cases. That structure leaves many everyday tools legal, but it adds disclosure duties where the risk is deception, impersonation, manipulation or hidden automated classification. (digital-strategy.ec.europa.eu) (digital-strategy.ec.europa.eu) The next deadline is now close enough that European product teams have to turn policy text into interface text. By August 2, 2026, “AI-generated” and “you are interacting with AI” need to work as operational features, not just compliance promises. (digital-strategy.ec.europa.eu) (digital-strategy.ec.europa.eu)