Berliner Tageblatt - Microsoft to invest $3.2 bn in AI in Sweden

The group will train 250,000 people by 2027 to boost AI knowledge and competence and also increase capacity at its three data centres in the country, it said.

"Microsoft's largest investment in our history in Sweden" would enable the Scandinavian country "to build world-leading AI data centre infrastructure," the company's president and vice chair Brad Smith said at a press conference with Swedish Prime Minister Ulf Kristersson.

"A big part of the reason we're able to do this is because of Sweden's forward-looking energy policy, the plentiful access to green energy, whether it's carbon-free energy or renewable energy," Smith said.

The US group has in recent months announced similar AI investments in other countries, including in France where it vowed to invest four billion euros ($4.3 billion), Japan where it announced a $2.9 billion AI push, and in Indonesia and Malaysia.

In Sweden, Microsoft will provide more than 20,000 graphic processing units (GPUs), needed for training AI models, and will boost capacity at its data centres in Sandviken, Gavle and Staffanstorp.

"AI is a catalyst for many things," Kristersson said. "It will also help accelerate development in other areas. This huge investment in Sweden has the potential to pave the way for other investments."

Data centres, which crunch and stock vast amounts of data, require large amounts of electricity and water, accounting for about two percent of global electricity consumption, according to a study by the HEC Montreal business school.

In 2020, Microsoft said it aimed to become "carbon negative" by 2030, but in 2023 its emissions rose by 30 percent, its data showed.

- 'Rise in Russian deepfakes' -

Asked about the risks and abuses associated with artificial intelligence, Smith said Microsoft was monitoring AI-generated deepfakes "very closely".

"Our biggest concern, to be honest, is about the Russian government," he said. "We've seen an increase in Russian activity using deepfakes."

"This is the kind of danger for the future that we need to address and protect against, and that's going to require more work."

He said it would require governments to introduce new legislation, as well as new capabilities in the tech sector.

"Fundamentally, it requires the use of AI to defend against abuses that others are advancing with AI," Smith said.

AI technology, which is expected to transform nearly every aspect of human life in the coming years, took a huge leap forward with the 2022 launch of the generative tool ChatGPT, which can create texts, images and audio files upon demand.

H.Seidel--BTB