By MATT O’BRIEN
An investigation by U.S. antitrust enforcement authorities is underway into the connections between major artificial intelligence (AI) companies like OpenAI, the creator of ChatGPT, and the tech giants that have heavily invested in them.
The probe seeks to capitalize on the AI boom driving the demand for chatbots like ChatGPT and other AI tools that generate original content. It specifically focuses on Amazon, Google, and Microsoft.
During an AI event on Thursday, Lina Khan, the head of the U.S. Federal Trade Commission (FTC), highlighted the scrutiny on whether these relationships allow dominant firms to wield undue influence or gain preferential access that could harm fair competition.
The investigation will delve into “the collaborations and partnerships forming between AI developers and major cloud service providers,” according to Khan.
The FTC has issued “compulsory orders” to five companies, including cloud service giants Amazon, Google, and Microsoft, as well as AI startups Anthropic and OpenAI, mandating disclosure of contract details and related decision-making processes.
One prominent partnership under review is Microsoft’s longstanding collaboration with OpenAI. Anthropic, a San Francisco-based AI firm founded by former OpenAI executives, recently struck multimillion-dollar deals with Google and Amazon.
While Google and Microsoft did not provide immediate comments, OpenAI, Anthropic, and Amazon declined to respond.
The link between Microsoft and OpenAI is also under scrutiny by the European Union and the UK. The EU’s executive branch announced an examination into potential antitrust issues arising from the partnership, which could impact merger regulations affecting competition within the EU. A similar inquiry was initiated by the UK’s competition authority in December.
Antitrust activists have welcomed the actions taken by the FTC and European regulators, characterizing these relationships as quasi-mergers that warrant intervention.
Matt Stoller, the research director at the American Economic Liberties Project, emphasized the need for regulatory intervention, stating, “Big Tech firms understand they cannot outright acquire top AI companies, so they seek to exert influence through other means.”
The total amount of Microsoft’s investment in OpenAI, described by Microsoft CEO Satya Nadella as a “complex arrangement,” has not been publicly disclosed.
Nadella hinted at a significant investment during a radio interview with IT journalist Kara Swisher in November, emphasizing that the support extended beyond financial contributions.
OpenAI faced internal turmoil when CEO Sam Altman was abruptly ousted by the board but later reinstated amidst widespread attention. The company underwent restructuring to stabilize operations, with Microsoft backing the majority of the board changes.
Despite lacking direct control, Microsoft secured a nonvoting board seat at OpenAI, aligning with the company’s unique governance model. OpenAI, initially established as a nonprofit research institute dedicated to safe AI development, now operates a for-profit division alongside its volunteer-based structure.
Prior to the launch of ChatGPT, Microsoft made a substantial $1 billion investment in OpenAI in 2019. The collaboration involved Microsoft providing computational resources for training large-scale AI models on extensive datasets, with resulting outputs integrated into Microsoft products.
Nadella likened the partnership to longstanding collaborations between Microsoft and Intel, emphasizing the mutual support between the two entities.
The FTC has been proactive in monitoring and preventing illicit activities related to AI technologies. Khan stressed the need to combat harmful practices involving AI-generated content, such as voice and image manipulation for fraudulent purposes.
Khan’s focus extends beyond immediate concerns to address the broader consolidation of power among a select group of AI leaders. The FTC aims to prevent anticompetitive practices that could stifle innovation and competition in the AI sector.
The FTC commissioners, all Democrats due to vacant seats, unanimously authorized the information requests to AI developers, akin to actions taken in previous investigations. Commissioner Alvaro Bedoya emphasized the importance of shedding light on the dynamics between these powerful AI models and the companies utilizing them.
The five companies have 45 days to disclose details of their partnerships, resource dependencies for AI development, and decision-making processes related to product launches.
Contributions to this article were made by Kelvin Chan, an AP business writer based in London.
