The U.K.’s antitrust regulator has launched a probe into Amazon.com Inc.’s partnership with large language model developer Anthropic PBC.

The Competition and Markets Authority, or CMA, disclosed the move today. The development is not entirely unexpected. In April, the CMA asked interested parties to submit feedback about the potential competitive impact of Anthropic’s work with Amazon.

The regulator stated in its brief announcement of the probe that it now has “sufficient information” about the partnership for an inquiry. The probe’s goal is to determine whether the companies’ collaboration should be referred to a Phase 2 antitrust investigation. Such an investigation could potentially lead the CMA to order that Amazon and Anthropic change certain terms of their partnership.

San Francisco-based Anthropic is backed by more than $7 billion in funding, $4 billion of which was provided by Amazon through a deal announced last September. The companies unveiled the investment alongside a broad cloud collaboration. Anthropic said at the time that it will rely on Amazon Web Services Inc. as its “primary cloud provider for mission critical workloads.”

The companies are working together in other areas well. Anthropic will help the cloud giant improve its Trainium and Inferentia chips for running artificial intelligence workloads, while Amazon will make the startup’s LLMs available to its engineers. AWS customers, in turn, also have access to Anthropic models through Amazon Bedrock.

When the CMA started collecting feedback for the probe in April, officials stated that the goal was to determine whether the partnership amounts to a so-called relevant merger situation. That’s a situation where two companies stop operating as separate businesses. If Amazon’s collaboration with Anthropic is found to meet the relevant criteria, the CMA will review whether the deal reduces market competition in the U.K. 

“We are an independent company,” Anthropic said in a statement responding to the probe. “Our strategic partnerships and investor relationships do not diminish our corporate governance independence or our freedom to partner with others.”

Anthropic’s other partners include AWS rival Google LLC, which last year committed to investing up to $2 billion in the LLM developer. The search giant earlier bought a stake worth more than $300 million. Anthropic has made its latest LLM, Claude 3.5 Sonnet, available on both AWS and Google Cloud.

An Amazon spokesperson told CNBC that “by investing in Anthropic, Amazon, along with other companies, is helping Anthropic expand choice and competition in this important technology. Amazon holds no board seat nor decision-making power at Anthropic, and Anthropic is free to work with any other provider (and indeed has multiple partners).”

Google’s partnership with Anthropic is likewise facing scrutiny from the CMA. Last month, the regulator began collecting feedback for a potential probe into the transaction.

Reports emerged last December that Anthropic was seeking to raise another $750 million in funding. Given the regulatory scrutiny of the company’s deals with Google and AWS, it may face challenges raising new capital from the cloud market’s largest players, at least in the near term. Google, AWS and Microsoft Corp. have provided a sizable percentage of the funding raised by generative AI startups over the past few years.

Anthropic reportedly targeted a valuation of up to $18.4 billion with the $750 million round it was rumored to be raising in December. In January, Forbes reported that the lead investor was expected to be not a major cloud provider but rather Menlo Ventures. 

Anthropic rival OpenAI is facing regulatory scrutiny as well. Last year, the CMA started collecting data about the ChatGPT developer’s partnership with Microsoft for a potential investigation. The deal has also drawn the attention of the European Commission, which recently signaled that it may launch an antitrust probe of its own.

Photo: Microsoft