Interviews with EU lawmakers and sources reveal how, over just 11 days, MEPs wrote a draft bill on generative AI, a rare example of quick consensus in Brussels
Context & Ripple Effects
The 11-day sprint by MEPs in spring 2023 was the opening move of a much longer legislative arc: that fast-written draft set the frame for regulating systems like ChatGPT before most capitals had a position. By December, the process hit its hardest test, with sources describing how lawmakers could not agree on how to regulate foundation models, which became the central sticking point of the AI Act talks.
That deadlock broke with the December deal on one of the world's first comprehensive attempts to limit AI use, including in law enforcement and critical infrastructure, followed by the member states' agreement pairing the Act with pro-innovation measures and a new AI Office. What this article adds is the origin story: the speed of the original consensus explains why the EU moved first at all.
First-order effects
- MEPs' rapid draft gave the EU a first-mover position on generative AI rules, meaning US-developed systems like ChatGPT faced a regulatory framework authored in Brussels before comparable legislation existed elsewhere.
Second-order effects
- The draft's scope pushed foundation models to the center of the negotiation, forcing model developers and EU governments into the standoff that stalled talks through late 2023 until member states traded compliance burdens for pro-innovation measures and an AI Office.
Third-order effects
- If the pattern holds, Brussels is using fast regulatory consensus as a tech-sovereignty instrument — cutting reliance on US tech giants by writing the rules they must operate under, a move EU businesses have flagged concerns about as compliance costs land on them too.
The trend: EU AI policy is evolving from a rapid-draft regulatory sprint into an institutionalized enforcement apparatus, with the AI Act doubling as Brussels' lever for reducing dependence on US tech giants.