Jaap Arriens | Nurphoto | Getty Images
The U.K.’s competition watchdog has opened an initial review into Microsoft’s mammoth investment into ChatGPT creator OpenAI, making it the first major regulator to flag potential competition concerns over the tech giant’s relationship with one of the most important artificial intelligence companies today.
The Competition and Markets Authority said in a statement Friday that it is seeking views from interested parties to address whether Microsoft’s $10 billion investment in OpenAI has led to a “relevant merger situation,” where two or more businesses have ceased or will cease to be distinct as a result of a transaction.
The CMA said the pace at which AI is scaling is “unrivalled in economic history,” and that advances in so-called foundation models, which describe general purpose AI tools such as ChatGPT, represent a “pivotal moment in the development of this transformative technology.”
The regulator said it will review whether Microsoft’s partnership with OpenAI has resulted in an acquisition of control – in other words, a situation where one company has material influence, de facto control, or more than 50% of the voting rights over another entity.
“The invitation to comment is the first part of the CMA’s information gathering process and comes in advance of launching any phase 1 investigation, which would only happen once the CMA has received the information it needs from the partnership parties,” Sorcha O’Carroll, senior director for mergers at the CMA, said in a statement.
Sam Altman’s exit — and return
The CMA is likely to have paid close attention to OpenAI CEO Sam Altman’s ouster from the board of the company he co-founded last month.
Altman was fired from OpenAI in a shock move from one of the companies at the heart of 2023’s boom in artificial intelligence.
The OpenAI board at the time said it “no longer has confidence” in Altman’s ability to continue leading OpenAI. Microsoft, which has an extensive partnership in place with OpenAI, subsequently hired Altman to lead a new advanced AI research team.
The move resulted in OpenAI employees revolting and calling for Altman’s return. In a matter of days, Altman returned to OpenAI as CEO and Microsoft obtained a board seat as a non-voting member.
Microsoft’s appointed board representative can attend OpenAI board meetings and access confidential information. But they don’t have voting rights on matters including electing or choosing directors.
Microsoft President and Vice Chair Brad Smith, responding to the CMA’s statement, called out Google’s 2014 acquisition of British AI lab DeepMind, saying that Microsoft’s partnership with OpenAI is unlike that deal.
“Since 2019, we’ve forged a partnership with OpenAI that has fostered more AI innovation and competition, while preserving independence for both companies,” Microsoft’s Smith said in an emailed statement to CNBC Friday.
“The only thing that has changed is that Microsoft will now have a non-voting observer on OpenAI’s Board, which is very different from an acquisition such as Google’s purchase of DeepMind in the UK. We will work closely with the CMA to provide all the information it needs.”
OpenAI said that Microsoft’s non-voting board seat does not conflict with the company’s independence.
“Our partnership with Microsoft empowers us to pursue our research and develop safe and beneficial AI tools for everyone, while remaining independent and operating competitively,” the OpenAI spokesperson told CNBC via email.
“Their non-voting board observer does not provide them with governing authority or control over OpenAI’s operations.”
Separately, the CMA is reviewing the AI industry to assess what risks and opportunities foundation models present, and what principles need to be applied to the tech to prevent competition and consumer protection breaches.
The European Union is also expected to agree on landmark rules designed to regulate AI technology soon.
The regulation, which has been in the works for years, has been the subject of tense negotiations by EU institutions, with disagreements mainly lying over how to govern foundation models and biometric identification tools.
