Transparency Rules Take Effect: AI Content in the EU Now Gets New Labels
The European Union has begun enforcing fresh transparency requirements for artificial intelligence as of August 2, based on Article 50 of the EU AI Act. Under these obligations, any content either generated or materially altered by AI must carry a label, and users must be told when they are dealing with a chatbot rather than a human. This is part of the broader AI Act, a landmark regulatory package being watched closely by businesses and policymakers around the world.
From now on, any material created with AI assistance—including realistic deepfakes, text, audio, or video—must be flagged. If the entire work was produced by AI, the label changes to “AI-Generated.” That requirement covers wholly AI-made fake videos of politicians, fabricated events, music, images, and news summaries.
Company Obligations and Penalties
Companies also have new duties: they must disclose when users are interacting with a chatbot or an AI agent instead of a person. Systems that recognize emotions or process biometric data need a separate label. Developers of general-purpose AI models must keep documentation, publish concise summaries of training data, and follow copyright policies.
Violating these rules can result in fines of up to €15 million or as much as 3% of a company’s worldwide annual revenue. EU institutions face a maximum fine of €750,000, while small and medium-sized businesses are covered by a proportionality principle. The rules do not apply to private correspondence, group chats, or works that are artistic, satirical, or fictional.
Engadget reported on the new requirements.
For an international audience watching tech regulation, this is a notable moment: the EU’s AI Act is a benchmark effort, and these transparency obligations are among its first concrete enforcement steps. Labeling AI-generated content is intended to guard against disinformation and promote honesty in how the technology is used. Other countries considering similar AI standards may look to these EU measures as a model.
As the EU implements these new transparency measures, tech giants are also adapting to the changing landscape. For instance, Google has announced plans to label advertisements that have been created or modified by artificial intelligence. This move highlights the growing emphasis on accountability and transparency in digital content, a trend that is likely to influence various sectors across the globe.