The 'AI war' has begun. Is Europe left empty-handed?

An existential European fear became reality last weekend when the US government ordered that the latest AI model from maker Anthropic be made inaccessible to non-Americans. The Trump administration cited advanced cybersecurity risks associated with the model, Fable 5, and its underlying language model, Mythos, as reasons for the restriction. Anthropic protested but was compelled by the government order to make Fable 5 globally inaccessible.

“

The AI war has begun.

This unprecedented, overt government intervention in artificial intelligence has generated significant international reactions over the past week. It is likely to have major consequences for AI development, the planned IPOs of Anthropic and other AI giants, and international relations. In Europe, reactions have focused primarily on the dangerous dependence on American technology. The feared "kill switch," a "red button" allowing the US president to shut down AI for other countries, appears to be real. French presidential candidate Gabriel Attal reacted sharply: "The AI war has begun."

“

A bizarre coincidence, but not entirely unexpected.

What should Europe do now? AI researchers Michiel Bakker (MIT, DeepMind) and Stan van Baarsen (Delta Institute) co-authored a future scenario, Europe 2031, depicting a Europe that is completely dependent and lagging behind by 2031. In their scenario, the US president uses a type of kill switch for the most advanced AI for Europe. They published their scenario one day before it actually happened. "A bizarre coincidence," says Van Baarsen by phone. "But hardly unexpected." The core of their concern is that AI is not just a handy app but crucial infrastructure for future prosperity, security, and progress. The most acute danger is cybersecurity, says Bakker: "If the Americans, with the help of advanced AI, can find all vulnerabilities in digital systems, but we cannot fix them because we do not have access to the same technology, then that can lead to a threat to our economy and society."

“

If the Americans, with the help of advanced AI, can find all vulnerabilities in digital systems, but we cannot fix them because we do not have access to the same technology, then that can lead to a threat to our economy and society.

The AI lag could also affect fundamental European innovation power, according to Bakker: "If scientists in the US soon have access to much better models than European scientists, you reach a point where all leading science, at least in the natural sciences, is driven by the Americans." Bakker, Van Baarsen, and their co-authors have received significant support from European tech entrepreneurs this week. They believe there is effectively no other choice but to join the AI race as the Americans are currently winning it: Europe, like the Americans, must invest hundreds of billions of euros in startups, existing companies, and the construction of huge data centers, even if they are initially very polluting.

“

If scientists in the US soon have access to much better models than European scientists, you reach a point where all leading science, at least in the natural sciences, is driven by the Americans.