The European Commission's AI Office has initiated enforcement of the EU AI Act by issuing information requests to over 30 general-purpose AI model providers. The office holds exclusive authority to regulate these models and systems integrated into large platforms, with the power to impose significant financial penalties for non-compliance.
This initial wave of information requests signals that the European Commission is prioritizing transparency in training data and copyright compliance for general-purpose models. For the streaming and advertising ecosystem, this centralized oversight means that AI-driven creative tools and recommendation engines used by platforms like Google and Meta must now meet strict machine-readable marking and safety standards. The office's ability to audit source code and model weights represents a significant shift in how tech giants must document their internal AI development processes. Industry observers should monitor the December 2, 2026 deadline, when existing generative systems must implement mandatory AI output labeling or face potential market withdrawal.
The EU AI Office's enforcement action arrives as the European Commission has already been testing its regulatory muscle under adjacent digital legislation. In February 2026, TikTok was charged with breaching EU online content rules over addictive features and told to change its app design or risk a fine of up to 6% of ByteDance's global turnover, establishing a pattern of simultaneous enforcement across multiple digital regulatory frameworks. By July 2026, the Commission preliminarily found Meta in breach of the Digital Services Act for the addictive design of Instagram and Facebook, focusing on features such as infinite scroll, autoplay, push notifications, and highly personalised recommender systems. This parallel pressure means that companies like Google, Meta, and Amazon now face compliance obligations under both the DSA and the AI Act simultaneously.
The enforcement timeline carries significant financial stakes for the named providers. Under the EU AI Act, penalties for non-compliance with general-purpose AI obligations can reach up to 3% of global annual turnover or €15 million, whichever is higher. The Commission's approach under the DSA provides a useful precedent for understanding how the AI Office may escalate. Politico reported that the Commission's TikTok case may set new design standards for the world's most popular apps, noting that the demand to disable infinite scrolling and change recommender systems does not bode well for other platforms under investigation. The AI Office's decision to issue information requests rather than formal investigations at this stage suggests a graduated approach, but providers that fail to respond adequately could face escalation to binding decisions and financial penalties before the end of 2026.
The scope of the information requests reflects the AI Office's focus on systemic risk assessment for frontier models. In April 2026, the Commission preliminarily found Meta's Instagram and Facebook in breach of the DSA for failing to prevent minors under 13 from accessing their services, with Executive Vice-President Henna Virkkunen stating that terms and conditions should not be mere written statements but the basis for concrete action to protect users. This enforcement philosophy appears to carry over into the AI Office's approach, where the 30 providers targeted by information requests must demonstrate not just policy commitments but verifiable technical compliance with transparency and safety obligations under the EU AI Act.
The European Commission's AI Office has officially launched EU AI Act enforcement by issuing formal information requests to over 30 general-purpose AI model providers. These inquiries focus on safety, security, and copyright transparency, marking a significant shift toward centralized oversight for large-scale AI systems and their impact on digital platforms.
The requests target safety, security, and copyright transparency for general-purpose AI models, ensuring providers meet strict machine-readable marking and safety standards.
Providers can face financial penalties of up to 3% of their worldwide annual turnover or €15 million, whichever is higher, for non-compliance or providing misleading information.
Providers must respond to the preliminary findings within 21 days, with a maximum submission length of 50 pages.
Existing generative systems must implement mandatory AI output labeling by December 2, 2026, or face potential market withdrawal.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source
