CMA Clears Microsoft-OpenAI Partnership: Implications for AI Competition

The Competition and Markets Authority (CMA), the U.K.’s competition regulator, recently decided that Microsoft’s partnership with OpenAI does not warrant an investigation under the merger provisions of the Enterprise Act 2002. This decision follows an investigation that began in December 2023, amid concerns that Microsoft’s influence over OpenAI could impact competition in the AI industry.

The CMA concluded that Microsoft does not control OpenAI’s commercial policy but instead exercises a significant level of material influence. The regulator found no evidence of a ‘change of control’ that would constitute a relevant merger situation. This conclusion was based on an evaluation of all available evidence, including Microsoft’s status as a major investor and its role in developing products like the Azure OpenAI Service, Copilot chatbot, and GitHub Copilot AI coding assistant.

Initially, the CMA was concerned that Microsoft’s control over OpenAI’s commercial policy had increased since 2019, potentially leading to a substantial lessening of competition in the U.K.’s AI sector. The agency feared that Microsoft could restrict rivals’ access to OpenAI’s advanced AI models, particularly given Microsoft’s strong market positions. There were also concerns about the partnership’s impact on the market for accelerated compute, with OpenAI potentially serving as a key customer.

However, recent developments have seemingly reduced Microsoft’s influence. In January, Microsoft restructured its cloud computing agreement with OpenAI, adopting a model where it retains the ‘first right of refusal’ for certain workloads. Microsoft also granted waivers for OpenAI to expand its computing capacity with other investors, such as SoftBank. Additionally, Microsoft abandoned plans to take a board seat at OpenAI, which could have triggered further regulatory scrutiny.

The CMA’s decision highlights the fluid nature of tech partnerships and the challenges regulators face in assessing control and influence. While the agency continues to scrutinize AI investments to prevent market consolidation, it has yet to find evidence of anticompetitive behavior. This case underscores the complexity of regulating rapidly evolving sectors like AI, where partnerships and collaborations are crucial for innovation.