European AI whirlwind week: Paris AI summit, Le Chat, AI Liability, and GPAI Code of Conduct
European AI Whirlwind Week: Paris AI Summit focused on business, JD Vance criticized Europe’s AI stance, EU pledged €200B, while the U.S. and UK refused to sign an AI declaration. Mistral AI launched Le Chat, the AI Liability Directive was dropped, and the GPAI Code sparked regulatory concerns.
All eyes were on Paris this week - the AI Action Summit in Paris marked a significant shift in focus compared to previous AI summits. The inaugural Bletchley AI Summit in the UK and the subsequent South Korea meeting centered on AI threats and security concerns. However, this week’s summit in Paris pivoted towards business, competition, and economic growth.
As expected, U.S. Vice President JD Vance delivered sharp remarks on Europe’s cautious approach to AI development, contrasting it with the entrepreneurial American model. The real test, however, lies ahead as we await the details of the new Executive Order on AI in the U.S. Insiders anticipate key decisions to be made within the next 5-6 months, determining whether this directive will be an improvement over its predecessors.
Meanwhile, the European Commission announced a €200 billion investment in AI, underscoring its commitment to strengthening AI development across the continent. Taking the stage, French President Emmanuel Macron seized the moment to highlight France’s nuclear energy capabilities, quipping “plug, baby, plug” - a subtle jab at former U.S. President Donald Trump’s “drill, baby, drill” ambition on fossil fuel extraction.
The U.S. and the UK were the only countries present that did not sign the joint Paris declaration on “inclusive and sustainable AI.” As the global AI race accelerates, the West still struggles to reach a consensus on harmonized rules and standards for AI, highlighting ongoing divisions in regulatory approaches.
Mistral AI Unveils ‘Le Chat’
A few days before the Paris AI Summit, European AI powerhouse Mistral AI unveiled its new chatbot, Le Chat, which is reportedly fully open-source, free, and 13 times faster than ChatGPT.
The announcement was met with pride and excitement across Europe, a stark contrast to the criticism Mistral previously faced for opposing certain AI Act provisions and later accepting Microsoft’s investment, reigniting debates over its sovereignty and reliability.
A break for AI Liability Directive
This week, the European Commission released its legislative calendar, with two notable omissions - the AI Liability Directive (AILD) and e-Privacy regulation.
AILD was originally intended to establish a legal framework for seeking compensation for AI-related damages. The withdrawal sparked mixed reactions: some welcomed the decision, while others criticized the lack of a harmonized EU-wide approach, leaving national courts to handle AI liability on their own.
A divide is also apparent within the Commission itself. While Executive Vice President Henna Virkkunen reportedly supported withdrawing AILD, Irish Commissioner Michael McGrath appears far less enthusiastic about the move.
AI regulation is not over: GPAI Code of Conduct incoming
For those who thought AI regulation debates ended with the AI Act and the AI Liability Directive, think again. The EU is currently working on the General-Purpose AI (GPAI) Code of Conduct, which EU Member States will decide on later this spring.
The official goal of the GPAI Code is to help companies adapt to the AI Act’s requirements coming in August 2025. However, critics argue that it serves as a backdoor for additional regulatory layers, incorporating elements that were left out of the AI Act, thereby creating further uncertainty.
Tech giants like Meta and Google have already voiced concerns over its implications.
A key question remains: how does the GPAI Code of Conduct align with Europe’s broader ambition to foster AI unicorns? While the GPAI Code of Conduct provides exemptions for SMEs, European current and future hyperscalers are likely to face yet another layer of compliance.
