European companies are preparing for new AI transparency rules under the EU AI Act, which take effect on August 2. The regulations require businesses to disclose when customers interact with AI systems and to label AI-generated content such as images, audio, and video. This marks the first consumer-facing obligations under the landmark legislation, with the European Commission empowered to impose fines for non-compliance.

Lawyers and industry experts compare the upcoming changes to the widespread adoption of cookie banners under EU privacy laws. Patrick Van Eecke, a lawyer at Cooley, noted that until now the AI act has been largely abstract, but soon consumers will see visible labels indicating AI involvement.

While the EU aims to boost AI adoption and competitiveness, companies like Zalando and Booking.com have expressed concerns over the clarity and timing of the new guidelines. Zalando criticized the Commission's guidance as adding unnecessary regulatory layers, while Booking.com highlighted the need for clearer implementation instructions despite supporting transparency.

The new rules are part of a broader rollout of the AI Act, with high-risk system requirements delayed until December 2027. The Commission can fine companies up to 3% of global annual revenue or €15 million for violations. Industry groups warn that overly broad definitions could lead to compliance fatigue and reduced innovation.