VIP Deep Dive: Inside the EU AI Act's First Compliance Wave - Who's Ready, Who's Not, and What's Coming in 2027

The EU AI Act's labeling rules are now in force. From this week, companies must clearly mark AI-generated images, audio, and text that is designed to look real. The next big compliance wave covers general-purpose AI (GPAI) providers and high-risk systems and arrives in August 2027. Most companies are not ready.

What the First Wave Required

The labeling rules are the first compliance milestone of the AI Act. They cover three categories: chatbots, deepfakes, and AI-generated marketing copy. Labels must be "clear, conspicuous, and machine-readable." The European Commission has published a template, and a dedicated helpdesk is operational for SMEs.

"AI labels are the new cookie banners. Annoying at first, then normalized," the FT wrote.

What Comes Next

The next wave of obligations lands in August 2027. GPAI providers will need to publish training data summaries, copyright compliance documentation, and detailed model evaluations. High-risk AI systems - covering employment, credit, education, and certain public-sector uses - will need conformity assessments, post-market monitoring, and human-oversight documentation. The compliance stack is much heavier than what labeling required.