Europe faces a stark choice on artificial intelligence: shun the technology and risk economic stagnation, or embrace it and risk dependency on US and Chinese developers. The continent's limited role in the global AI safety debate highlights its struggle to produce homegrown tech giants.
Europe's struggle to establish a strong presence in AI development and safety discussions risks economic disadvantage and technological dependency on the US and China, potentially impacting its sovereignty and ability to shape global AI governance.
Europe faces a significant challenge in the global artificial intelligence race, grappling with the dilemma of either shunning the technology and risking economic growth, or embracing it and becoming dependent on tools developed by the US and China. This dilemma was underscored by Christine Lagarde, head of the continent's central bank, who asserted that Europe must develop its own AI technology and build more datacenters to avoid being cut off by other global powers.
While Europe has been largely absent from the highest levels of the AI safety debate, a situation attributed to its inability to produce tech titans on the scale of Apple or Google, its regulatory prowess is noted. European Commission President Ursula von der Leyen announced plans to convene discussions with leading AI labs to explore ways to pace the development of the technology. However, her remarks did not garner the same attention as dismissals of a slowdown from Donald Trump and China, highlighting Europe's limited leverage without their support.
Margrethe Vestager, a former EU competition commissioner, co-authored a report echoing Lagarde's concerns, warning that Europe is already experiencing economic and social consequences from decisions made elsewhere. The report suggests building more datacenters and ensuring resilience against potential AI crises. Vestager also pointed to the EU AI Act as a crucial piece of legislation that mandates safety incident reporting and risk mitigation for high-risk AI systems. However, she acknowledged that existential threats require global solutions that do not cede power to the US and China.
Frederike Kaltheuner, an adviser to the AI Now Institute, suggests that Europe's dependency on US technology is substantial and difficult to reverse. She argues that the narrative of Europe lagging significantly is pushed by American tech executives to promote their own interests. Kaltheuner notes that many companies are hesitant to adopt expensive "frontier models" due to cost and data concerns. If this trend persists, Europe's position might not be as precarious as portrayed, especially with the EU AI Act's pragmatic measures for consumer-facing AI.
Itxaso Dominguez de Olazabal, a policy adviser at EDRI, agrees that Europe has a competitiveness problem but criticizes US tech proponents for blaming regulations and promising deregulation as a solution. Georgina Kon, a partner at Linklaters, expressed doubt that the EU AI Act will become a global benchmark due to its compliance burden and limited territorial scope, suggesting that other countries are adopting different legislative approaches. Despite these challenges, big tech maintains a significant customer base in Europe, and Kaltheuner worries that doomsday narratives could push Europe towards dependency on US-controlled models and infrastructure, posing a sovereignty problem.