The EU’s landmark AI Act enters a new phase this week as regulators begin enforcing rules requiring companies to label AI-generated content and disclose when users are interacting with technology. From August 2, the European Commission's AI Office, alongside national authorities, started enforcing key provisions, including transparency obligations for general-purpose AI models and bans on prohibited practices.

Under these rules, businesses must clearly inform users when they are interacting with AI rather than a human. AI-generated or manipulated content, such as deepfakes, must be labeled, with certain content requiring machine-readable markers. The AI Office will oversee providers of general-purpose AI models, while national regulators enforce rules for most other AI systems.

Companies found in breach could face fines of up to 3% of their annual global turnover. This move is causing significant concern for US AI labs, as any lab with European customers is considered in scope. While regulators can levy substantial penalties, enforcement is expected to build gradually as companies demonstrate they are working towards compliance.