EU AI Act Enters Second Wave of Enforcement Actions
Brussels has opened investigations into three major model providers over transparency and copyright disclosure obligations under the AI Act.
W
We Read AI
We Read AI
The European Commission has opened investigations into three major AI providers over alleged non-compliance with the AI Act's transparency and copyright disclosure requirements. This is the second wave of formal enforcement since the Act's key provisions took effect.
Who's under scrutiny
The Commission has not named the targets publicly, but industry reporting points to two US frontier labs and one European provider. The specific claims involve incomplete training data disclosure and insufficient copyright-use documentation.
What's at stake
Penalties under the AI Act can reach 3% of global annual turnover — meaningful sums for providers with multi-billion revenue runs. More importantly, formal findings would set precedent for compliance patterns across the industry.
The compliance response
Legal teams at major labs have accelerated the buildout of "training data manifests" — structured logs of what a model was trained on and how disputed material was handled. Expect these to become standard artifacts of new model releases.
Source: [European Commission press release]
#eu#regulation#ai-act#copyright