Microsoft and Google have invested in AI in Europe

Big tech investments may be aiming to rebuild their AI infrastructure ahead of new legislation from Brussels.

Google's parent companies Alphabet and Microsoft have both announced huge investments in artificial intelligence in Western Europe this week, likely ahead of EU AI regulations that will be introduced in April.

 

According to the French finance ministry, a new Google AI hub in Paris will eventually host more than 300 Google researchers and engineers, including members of the Chrome and YouTube development teams.

 

The new AI hub aims to accelerate the development of AI-based products in France, create new academic and research partnerships, and increase the adoption of AI tools among French workers, the French finance ministry said in a February 14 news release. release

 

France has promoted itself as a European hub for AI development. For example, Mistal AI, a French AI company, was founded in April 2023 by researchers who previously worked for Google DeepMind and Meta. The company already has a valuation of around US$2 billion, having raised around US$415 million in an initial funding round

 

Meanwhile, Microsoft will invest €3.2 billion (US $3.4 billion) by the end of 2025 to build the company's AI infrastructure and cloud capabilities in Germany.

 

"Microsoft is committed to enabling companies across the German economy to use AI to stay ahead of global competition," Microsoft President Brad Smith wrote in a tweet. "This... investment will double Microsoft's AI infrastructure and cloud capacity to meet the country's accelerating demand for AI services."

 

German Chancellor Olaf Scholz praised the investment, saying in a news release that it will improve computing infrastructure in the county and strengthen the German AI ecosystem.

 

According to Smith, the investment will be used to build new data centers near Frankfurt and in North Rhine-Westphalia. Some of the funds will also go towards training some 1.2 million German workers.

 

Upcoming regulations

 

Both companies' investments in Western Europe appear to be an effort to comply with EU AI regulations, said Gartner senior vice president analyst Jason Wong. “Microsoft, Google, and other companies that host generative AI models need to build the necessary infrastructure. And operations in Europe must comply with existing and emerging regulations on AI data and use by consumers and businesses,” he said via email.

 

This week, two groups of lawmakers in the European Parliament approved a tentative deal as a step toward an April vote on landmark legislation that would create the world's first AI regulation. The proposed EU AI law would ban the use of AI to create "unacceptable risks", including the cognitive behavioral manipulation of people or the categorization of people based on behavior, socio-economic status, or other personal characteristics.

 

The law will also require AI tools representing "high risk" to be registered in the EU. High-risk AI applications include cars, medical devices, and aviation AI developers must register AI used in law enforcement, education, critical infrastructure management, and many other applications.

 

Common-use AIs such as ChatGPT or Google's Gemini will be subject to transparency requirements, including the disclosure of AI-generated content. EU law requires AI developers to design their models so that they cannot create illegal content.