The EU AI Act entered full enforcement on August 2, 2026, activating transparency obligations under Article 50, GPAI enforcement powers, and penalty regimes for prohibited and high-risk AI systems. [1][2][3][4][5][6][7]
Enforcement has ramped up with information requests sent to 30+ companies including Anthropic and OpenAI, targeting cybersecurity, safety, and copyright practices, following AI-related incidents. [8][9]
The European Commission confirmed the AI Act is fully enforced and no longer just rules on paper, with penalties for prohibited AI applications reaching €35 million or 7% of global annual turnover and fines up to €15M/3% turnover for GPAI violations. [10][11][12][13][14]
The EU AI Office is hiring 40 additional enforcement staff to support the enforcement ramp-up, with applications closed on September 8, 2026. [15][16]
Companies like Anthropic have detailed text watermarking systems based on SynthID-Text for EU AI Act transparency compliance, reflecting ongoing adaptation to Article 50 obligations. [17]
No formal enforcement action against a specific company was announced for September 4–6, 2026, indicating a phase of information gathering and preparation. [18]