In the span of 48 hours, the European Parliament reportedly reached a preliminary agreement on high-risk compliance deadline extensions, the European Council proposed a prohibition on AI-generated non-consensual intimate imagery, and the European Commission's second draft Code of Practice on AI content labeling opened for feedback with a reported March 30 deadline. These three developments are moving on separate tracks, but they converge on a single question for compliance teams: what's changing, what isn't, and what should you do about it right now?
Three EU AI Act developments. Three different EU institutions. One 48-hour window.
This kind of convergence isn’t coincidence, it’s the EU legislative machinery moving toward a final shape of the AI Act and its implementation framework. But the three tracks are distinct, and the practical implications for compliance teams differ significantly across them. Treating them as a single story would be a mistake.
Here’s the breakdown, and a triage guide for what to prioritize.
