Europeans Are About to Find Out How Entrenched AI Is in Their Daily Lives
New EU rules stipulate that people must be told when they’re interacting with AI or looking at AI-generated or -edited content, leading to fear of “disclosure fatigue.”
On August 2, the European Union will enforce new transparency rules requiring businesses to disclose when they are interacting with artificial intelligence systems or viewing content produced or modified by the technology. This is part of the bloc’s broader Artificial Intelligence Act, designed to curb deception and manipulation while promoting trustworthy AI in Europe.
Frederiek Fernhout, a technology lawyer at Stibbe, observes that compliant providers will make it clear how extensively AI is utilized, particularly in marketing materials. Ads or corporate social media posts generated using deepfakes will bear a label indicating AI generation. Chatbots and complaint handling hotlines must explicitly state their AI-based nature, while call centers leveraging machine learning to monitor emotional cues and detect frustration will need to disclose this capability at the start of the call.
Businesses are also obligated to declare any commercial use of AI technology, including tasks like scheduling appointments, managing correspondence, or negotiating contracts. Non-compliance could result in substantial fines of up to €15 million ($17 million) or 3 percent of the company’s global annual turnover, whichever is greater.
The legislation extends to model developers, who will be monitored by the European Commission’s newly formed AI Office. This encompasses companies beyond the well-known names like OpenAI, Anthropic, and Google DeepMind, extending to others such as Spotify and Adobe, which utilize AI-powered recommendation engines and editing features in Photoshop, respectively.
While critics argue that excessive labeling could lead to information overload and desensitize users ("banner blindness"), the law echoes the EU’s 2018 General Data Protection Regulation (GDPR), which granted consumers greater control over personal data and introduced ubiquitous cookie consent notifications, leading to "cookie fatigue."
Fernhout notes that several terms are unclear and require specific guidance from authorities, similar to GDPR, emphasizing the technical nature of these regulations and the need for built-in compliance within systems. A transitional period until December allows AI model providers to label synthetic audio, images, video, or text in a machine-readable format, acknowledging the technical challenges of implementation. However, the European AI Office’s delayed production of key compliance guidelines has delayed enforcement.
With 27 member countries at varying stages of establishing supervisory frameworks, enforcement may not be immediate or uniform across the EU, warns Rosie Nance, a data and AI regulatory lawyer at Norton Rose Fulbright. Despite these challenges, the legislation is expected to be a transformative shift for companies, requiring them to decide whether to label AI usage or revert to non-AI-based systems.
Written by urgent.news from Wired Business's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
