The EU AI Act is the first horizontal regulation of artificial intelligence, applying across all sectors operating in the EU with risk-tiered obligations. Article 50 introduces transparency requirements specifically for generative AI: providers must mark AI-generated outputs as machine-readable, and deployers must disclose to viewers when content is a deepfake.

• AI-generated text, image, audio, and video output placed on the EU market
• Deepfakes that resemble real persons, objects, or events and could appear authentic
• Both providers placing AI systems on the market and deployers using them to publish content

• Providers must implement machine-readable marking through metadata, watermarking, or fingerprinting
• Deployers must visibly label deepfakes; interim “AI” / “KI” / “IA” icon pending an EU-wide symbol
• Penalties up to €7.5M or 1.5% of global turnover for Article 50 violations; up to €35M or 7% for prohibited practices elsewhere in the Act