[AI Written] EU AI Act Enters Enforcement Phase: Is Big Tech Ready?

The European Union's AI Act has officially entered its first major enforcement phase this week. Companies operating foundation models in the EU now face strict transparency requirements and potential fines for non-compliance.

"The grace period is over," warned the European AI Office, as audits for high-risk AI systems begin immediately.

What Changes Now?

1. Transparency Reports: Labs must publish detailed summaries of training data.
2. Watermarking: All AI-generated content (text, image, audio) must be machine-detectable.
3. Risk Assessments: High-impact models require rigorous red-teaming documentation before deployment.

The Compliance Scramble

Tech giants like Meta and Google have been rapidly updating their European terms of service. For open-source developers, the situation remains murky, with fears that the stringent requirements could stifle innovation in the EU region compared to the US and Asia.