In this context, Europe’s reliance on American-made AI is starting to look more and more like a burden. In a worst-case scenario, although experts consider this a remote possibility, the United States could decide to withhold access to artificial intelligence services and key digital infrastructure. It is more likely that the Trump administration could employ Europe’s dependence as leverage if both sides continue to do so conclude a trade agreement. “This dependence is a burden in any negotiations, and we will negotiate with the US more and more often,” says Taddeo.
The European Commission, the White House and the UK Department of Science, Innovation and Technology did not respond to requests for comment.
To hedge against this risk, European countries have tried to move AI production offshore financing programs, targeted deregulationand partnerships with academic institutions. Some efforts have focused on building competitive vast language models for native European languages, e.g Open AND GPT-NL.
But as long as ChatGPT or Claude continue to outperform European-made chatbots, America’s AI advantage will only grow. “In these domains, often the winner takes all. When you have a very good platform, everyone goes there,” says Nejdl. “Not being able to develop cutting-edge technology in this field means you won’t catch up. You’ll always just feed the bigger players with your input, making them even better, and leaving you further behind.”
Watch out for the gap
It is unclear how far the UK and EU intend to go in pushing for “digital sovereignty” – say lobbyists. Does sovereignty require complete self-sufficiency in a extensive AI supply chain, or just better capabilities in a narrow set of disciplines? Does it require the exclusion of US-based suppliers or only the availability of domestic alternatives? “It’s quite unclear,” says Boniface de Champris, senior policy manager at the Computer & Communications Industry Association, a membership organization of technology companies. “It seems to be more of a narrative at this stage.”
There is also no broad agreement on what policy levers should be used to create the conditions for Europe to become self-sufficient. Some European suppliers are advocating a strategy whereby European companies would be required or at least encouraged to buy from domestic AI companies – similar to China’s reported approach to its domestic processor market. Unlike grants and subsidies, this approach would aid seed demand, argues Ying Cao, chief technology officer at Magics Technologies, a Belgian company developing AI-specific processors for employ in space. “This is more important than simply access to capital,” Cao says. “The most important thing is that you can sell your products.” But proponents of open markets and deregulation say trying to eliminate U.S. AI companies risks putting domestic companies at a disadvantage compared to global competitors who have the ability to choose the best-suited AI products for them. “From our point of view, sovereignty means the ability to choose,” says de Champris.
Yet, for all the disagreements over policy details, there is widespread belief that closing the performance gap with U.S. leaders remains entirely possible even for labs with confined budgets and resources, as DeepSeek illustrates. “If I thought right now that we wouldn’t catch up, I wouldn’t do it [try]”, says Nejdl. SOOFI, the open source model development project in which Nejdl is involved, intends to release a competitive general-purpose language model with about 100 billion parameters within the next year.
“Progress in this area will no longer depend largely on the largest GPU clusters,” says Nejdl. “We will be Europe’s DeepSeek.”