EU rolls back AI Act timelines

Ah, the EU AI Act — that magnificent piece of legislation that was going to save us from the machines — has been quietly shunted into the long grass. The timelines have been pushed back to 2027 and 2028, which is a neat trick when you consider that frontier models are doubling every few months. 'Simplification,' they call it, which in Brussels-speak means 'we couldn't agree on anything, so let's kick the can down the road until after the next Commission term.' One imagines the tech lobbyists are already uncorking the champagne, because nothing says 'responsible innovation' like a regulatory framework that arrives just in time to govern the AI of three generations ago.

The irony is almost too perfect: the very Act designed to keep pace with AI is now explicitly admitting it can't. The EU's own impact assessment warned that even a single year of delay could lead to significant regulatory gaps, but apparently the risk of upsetting Big Tech's business models was more pressing. So instead of binding rules by 2025 or 2026, we get a leisurely rollout that ensures the first major compliance deadlines will fall somewhere around the same time as the next ice age. 'Europe leads on regulation' has become 'Europe leads on rearranging the deckchairs.'

And let's not forget the real-world victims of this postponement. Every month without enforceable rules means more high-risk AI systems deployed without oversight — from hiring algorithms that discriminate to medical devices that hallucinate. But hey, the 'simplification' process needs time. Perhaps by 2028, the chatbots will have evolved enough to write their own compliance reports. Efficiency!