Starting Sunday, August 2, 2026, citizens of the European Union have the right to know when they are interacting with an artificial intelligence system or viewing content generated or altered by the technology. These new transparency obligations, introduced by the Artificial Intelligence Act, represent a groundbreaking shift in how companies and digital service providers must communicate with the public. The stated goal of the regulation is to reduce deception and manipulation, helping people make informed choices within the broader aim of fostering trustworthy AI in Europe.
The scope of the law is broad and touches many sectors. Advertisements and social media posts using deepfakes must be labeled as AI-generated. Chatbots and hotlines handling complaints or customer support must clearly state if they are AI-based. Call centers using machine learning to monitor caller emotions and detect frustration must inform users at the start of the call. But it doesn't stop there. Any commercial use of the technology, from scheduling appointments to handling correspondence to negotiating contracts, must be declared. Penalties for non-compliance are severe, with fines up to €15 million ($17 million) or 3% of worldwide annual turnover, whichever is higher.
Sponsored Protocol
The regulation extends to models and tech companies
The restrictions are not limited to end users. AI model developers will also be subject to the new law, overseen by the European Commission's recently established AI Office. This includes not only giants like OpenAI, Anthropic, and Google DeepMind, but also companies not always directly associated with AI. Thibau Duquin, a technology and data lawyer at Stibbe, points to Spotify, which uses AI-powered recommendations, and Adobe for its AI editing features in Photoshop. "When you start looking, it's everywhere, and all this content will have to be labeled from now on," says Duquin.
The risk of label fatigue and comparison with GDPR
Critics are not hard to find. Excessive labeling could lead to a sort of "banner blindness," with endless notifications losing their meaning. Boniface de Champris, AI policy lead at the Computer & Communications Industry Association, warned that if everything is labeled, from spell-checked emails to filtered photos, AI labeling becomes meaningless. The new regulation echoes the GDPR, the privacy law that came into force in 2018, which introduced ubiquitous cookie-consent pop-ups and led to "cookie fatigue." In that case, technical challenges and interpretation also caused initial difficulties. Frederiek Fernhout, a technology lawyer at Stibbe, stresses that many terms are unclear and require specific guidance from authorities. Moreover, the requirements are highly technical and must be built into underlying systems, further complicating implementation.
Sponsored Protocol
Transition period and non-uniform enforcement
The EU has allowed a transition period until December for AI model providers to label synthetic audio, image, video, or text in a machine-readable format so that this material can be detected as AI-generated, clearly recognizing the technical challenge. Other compliance deadlines have been delayed because the European AI Office has taken longer than expected to produce key guidelines. The 27 member states are also at different stages of setting up supervisory frameworks, meaning enforcement might not be immediate or uniform. Rosie Nance, a data and AI regulatory lawyer at Norton Rose Fulbright, cautions that we won't necessarily see consistent enforcement right away.
Sponsored Protocol
Despite these challenges, the sweeping reach of the regulation will be a "gamechanger" for companies, says Duquin. Businesses will have to decide whether they are willing to label their AI use or revert to manual methods. Either way, transparency will become a central element in the relationship between consumers and technology. The use of AI in real-world scenarios is already pervasive, as shown by its adoption in US fast food, and this European regulation marks a significant step for public awareness.
