EU AI Act enforcement begins with bans on social scoring, biometric surveillance

The European Union began enforcing the first tranche of its AI Act on August 2, 2026, targeting the most dangerous AI applications and requiring transparency from general-purpose models. The rules apply to any organization that deploys or develops AI systems within the EU, regardless of where the company is headquartered.

Eight categories of AI practices are now banned outright, including social scoring by governments, real-time biometric surveillance in public spaces, and AI systems that exploit vulnerabilities of children or other vulnerable groups. Companies that violate these prohibitions face fines of up to 7% of global annual turnover or 35 million euros, whichever is higher.

Transparency rules take effect for chatbots and deepfakes

AI systems that interact with humans, such as chatbots, must now disclose that users are speaking to a machine. Deepfakes and AI-generated content must be clearly labeled. General-purpose AI models like those from OpenAI and Google must publish summaries of their training data and comply with EU copyright law.

• Providers of prohibited AI systems must cease operations immediately or face enforcement action.
• Developers of general-purpose AI models must register with the EU AI Office and provide documentation on model capabilities and limitations.
• Companies using AI for employee monitoring or credit scoring face additional obligations under the high-risk category, though those rules do not take effect until 2027.
• National market surveillance authorities in each member state are now empowered to conduct audits, request information, and impose fines for non-compliance.

Phased enforcement continues through 2027

The next compliance deadline is August 2027 for high-risk AI systems used in critical infrastructure, education, employment, law enforcement, and access to essential services. The European Commission has established the AI Office to oversee implementation and coordinate with national regulators. Industry groups have raised concerns about regulatory burden and potential fragmentation, but consumer advocates welcome the protections as a global benchmark. The enforcement marks a significant step in AI regulation, with other jurisdictions including the United States and Japan closely monitoring the EU's approach.