EU Gains New AI Act Enforcement Powers as Transparency Rules Take Effect

The European Union's third wave of AI Act requirements came into force on August 2, giving the European Commission's AI Office full investigatory and enforcement powers over general-purpose AI models and activating Article 50 transparency obligations for AI systems, CNBC and Al Jazeera reported this week.

New Powers, New Fines

Under the rules, the Commission can demand evaluations of models before public release in the region, restrict EU market access, and fine a provider up to 15 million euros ($17.3 million) or 3% of global annual turnover, whichever is higher. The powers raise the stakes for US companies including Anthropic and OpenAI, which have faced months of tension with Brussels over model access — most recently after cyber incidents involving their models, which Reuters reported the EU is now discussing with both labs.

Transparency: Chatbots Must Say They Are AI

Article 50 obliges chatbots and other interactive AI systems to tell users they are dealing with AI rather than a human, requires deepfakes — images, video, or audio generated or altered by AI — to be labeled, and mandates machine-readable marking so AI-generated content can be detected and traced. The Commission published guidelines on August 5 and a list of more than 180 organizations that have signed its Code of Practice on transparency of AI-generated content.

The High-Risk Delay

The Act's most demanding obligations — for high-risk AI systems in biometrics, employment, education, and migration — were pushed to December 2, 2027 as part of the Digital Omnibus package adopted in July, leaving some of the systems most exposed to automated decision-making without the Act's strongest protections for another 16 months. Digital rights groups have challenged the delay, warning it rewards industry lobbying and could set a precedent for further postponements.