This concentration of power is inconvenient for European governments. This will make European businesses the customers of the future, importing the latest services and technologies in exchange for money and data sent west across the Atlantic. These concerns have taken on a modern urgency – partly because some in Brussels see a growing divergence in values and beliefs between Silicon Valley and the average EU citizen and their elected representatives; and partly because artificial intelligence holds a gigantic place in the collective imagination as the driving force behind the next technological revolution.
European concerns about AI delays predated ChatGPT. In 2018, the European Commission released the AI plan calling for “European-made AI” that could compete with the US and China. However, beyond the desire to have some kind of control over the shape of the technology, the operational definition of AI sovereignty has become quite unclear. “For some, this means that we need to unite to oppose Big Tech,” says Daniel Mügge, a professor of political arithmetic at the University of Amsterdam who studies technology policy in the EU. “For others it means there is nothing wrong with Big Tech as long as it is European, so let’s crack it and make it happen.”
These competing priorities have begun to complicate EU regulations. The bloc’s law on artificial intelligence, which was adopted by the European Parliament in March and is likely to go into effect this summer, focuses primarily on regulating the potential harms and privacy issues associated with the technology. However, some member states, notably France, made clear during negotiations on the bill that they feared the regulation could hamstring nascent artificial intelligence companies that they hope will become a European alternative to OpenAI.
Speaking ahead of the UK’s November AI Security Summit, French Finance Minister Bruno Le Maire he said that Europe needs to “innovate before it regulates” and that the continent needs “European actors mastering artificial intelligence.” The final text of the Artificial Intelligence Bill includes a commitment to make the EU “a leader in the deployment of trustworthy AI”.
“At the last minute, the Italians, the Germans and the French thought, ‘Well, we need to cut the European companies some slack on the entry-level models,’” says Mügge. “This is based on the belief that Europe needs European artificial intelligence. “I feel like people have since realized that it’s a little bit more difficult than they would like.”
Sarlin, who recently traveled to European capitals, including: met with policymakers in Brussels, says Europe does indeed have some of the elements it needs to compete. To become an AI player, he says, you need data, computing power, talent and capital.
Data is quite widely available, adds Sarlin, and Europe has AI talent, although it sometimes struggles to retain it.
To mobilize more computing power, the EU is investing in high-performance computing resources, building a pan-European network of high-performance computing facilities and offering start-ups access to supercomputers through “AI factories” initiative.
Accessing the capital needed to build gigantic AI projects and companies is also a challenge, due to the gigantic gap between the United States and the rest of the country. According to Stanford University AI Index Reportprivate investment in US AI companies reached $67 billion in 2023, more than 35 times the amount invested in Germany or France. Research conducted by Accel Partners shows that in 2023, the seven largest private investment rounds by U.S. generative AI companies totaled $14 billion. The top seven in Europe raised a total of less than $1 billion.