Microsoft39s multi billion dollar investment in OpenAI could face EU merger

Microsoft's multi-billion dollar investment in OpenAI could face EU merger probe –

Jaap Arriens | Photo only | Getty Images

As part of its assessment, the Commission said it wanted to understand how competitive these markets currently are and to gain insight into how competition law could help these areas.

The EU Commission also said it was reviewing “some of the agreements concluded between major digital market players and developers and providers of generative AI” and highlighted the Microsoft-OpenAI connection as a particular deal it will investigate.

“The European Commission is examining whether Microsoft's investment in OpenAI may be reviewable under the EU Merger Regulation,” the Commission said in a statement on Tuesday.

Microsoft did not immediately respond to a CNBC request for comment.

The commission said it had sent requests for information to “several major digital players” and was also seeking views from interested parties, who have until March 11 to submit responses.

“Virtual worlds and generative AI are evolving rapidly,” Vestager said in a statement on Tuesday. “It is vital that these new markets remain competitive and that nothing stands in the way of businesses growing and providing consumers with the best and most innovative products.”

Microsoft has poured billions of dollars into OpenAI, the company behind the popular AI chatbot ChatGPT. The company has integrated some OpenAI technologies into its Office, Bing and Windows products and provides OpenAI with its own Azure cloud computing tools.

The Redmond, Washington-based tech giant first invested in OpenAI in 2019, contributing $1 billion in cash.

Then last year, the company made headlines when it reportedly invested an additional $10 billion in OpenAI, bringing its total investment to date to $13 billion.

News of the EU review comes after Britain's Competition and Markets Authority announced in December that it would launch an initial review of Microsoft's investment. The US Federal Trade Commission is also reportedly reviewing the merger, according to Bloomberg News.

The CMA said it was examining whether Microsoft's involvement in AI had created a “relevant merger situation”, citing “a number of developments in OpenAI's governance, some of which affected Microsoft” as a key cause for concern.

Earlier this year, OpenAI experienced a period of turmoil when its CEO Sam Altman was shockingly ousted from the board. In a dramatic turn of events, former Twitch CEO Emmett Shear briefly took over the top job before being ousted and replaced by Altman amid a reshuffle of the board.

As part of this update, Microsoft added its own observer to the board, raising concerns that the company may be exerting control over OpenAI. For its part, Microsoft has tried to emphasize that its designated board observer, recently named Dee Templeton, is a non-voting member.

At the time of the CMA announcement, Microsoft stated that it did not own any shares in OpenAI, while OpenAI stated that it remained independent and operated competitively.

“Their non-voting board observer does not give them any managerial authority or control over OpenAI’s operations,” an OpenAI spokesperson told CNBC via email in December.

“While details of our agreement remain confidential, it is important to note that Microsoft does not own an interest in OpenAI and is only entitled to a share in the distribution of profits,” Microsoft Chief Communications Officer Frank Shaw said in a December statement.

At the heart of the concerns is Microsoft's close partnership and investment in OpenAI, which gives the Redmond titan access to one of the most advanced developers of AI tools today. OpenAI's GPT-4 large language model can reportedly handle 25,000 words of input text – a big step up from the existing ChatGPT limit of 4,096 characters.

Scientists have said that GPT-4 has demonstrated human-level performance in various academic exams, while some AI researchers and technologists have speculated that GPT-4 could come close to artificial general intelligence (AGI), which could be as intelligent or more intelligent should as people.

—CNBC's Hayden Field and Kif Leswing contributed to this report