EU begins enforcing AI Act transparency rules as companies race to label chatbot and deepfake use

EU begins enforcing AI Act transparency rules as companies race to label chatbot and deepfake use

New obligations from August 2 require disclosure for many AI interactions and synthetic media, while the European Commission gains stronger oversight of general-purpose AI models.

Fact Check
Official European Commission sources directly confirm the claim. The 'AI Act | Shaping Europe's digital future' page and the 'Guidelines on transparency obligations' page both state Article 50 transparency rules apply from 2 August 2026, requiring disclosure when users interact with AI (chatbots) and marking/labeling of AI-generated synthetic media and deepfakes. Both official sources and the crypto news report confirm the AI Act reaches general application on 2 August 2026, when the AI Office/Commission and national authorities begin enforcement, giving the Commission stronger oversight of general-purpose AI models. The only minor nuance is that GPAI substantive obligations began in Aug 2025, but the Commission's direct enforcement powers over GPAI apply from Aug 2026, consistent with the claim's 'stronger oversight' framing.
Summary

The European Union began enforcing new AI Act transparency rules on August 2, extending disclosure requirements to chatbots, deepfakes and other AI-generated content used in the bloc, including by companies based outside the EU. AI systems that interact directly with people must make clear when users are dealing with AI, while providers of generative AI systems must add machine-readable identifiers to AI-generated or manipulated audio, images, video and text. The new phase also requires public distributors of realistic deepfakes depicting real people or events to clearly label the content as AI-generated or manipulated. The European Commission's AI Office and member state authorities will share oversight depending on their mandates, while the Commission now has immediate authority to supervise providers of general-purpose AI models and impose additional risk-management obligations on models deemed to pose systemic risks. Businesses across retail, travel, advertising and customer service are adjusting websites, hotlines and marketing workflows to comply. Lawyers said the rules could make AI use far more visible across commercial operations, from complaint-handling chatbots to sentiment analysis in call centers and AI-assisted advertising. Violations of chatbot and deepfake transparency obligations can draw fines of up to 15 million euros or 3% of worldwide annual revenue from the preceding fiscal year, whichever is higher. Companies and industry groups have warned about legal uncertainty, inconsistent enforcement across the EU's 27 member states and the risk of "label fatigue" if consumers are inundated with AI disclosures. The EU has said routine photo editing may fall under exceptions, and providers of AI models have until December to implement machine-readable labeling for synthetic media. The rollout is being compared with the GDPR's broad impact on digital business practices, and companies offering AI-powered services in the EU are now under pressure to determine how widely the rules apply to their products and content.

Terms & Concepts
  • deepfakes: Realistic audio, images or video that have been generated or altered using AI to depict people or events.
  • general-purpose AI models: AI models built for many different tasks and applications rather than one specific use.
  • systemic risks: Broader risks from powerful AI models that could have significant effects across society or markets.