The U.K.s’ antitrust authority has concluded that Amazon’s partnership and equity investment in AI startup Anthropic can’t be investigated under current merger rules due to the size and scope of the deal.
The U.K. Competition and Markets Authority’s (CMA) announcement comes six months to the day after news emerged that Amazon had completed a $4 billion investment in Anthropic, one of several heavily funded AI-focused startups. The three-year-old firm develops large language models (LLMs) and an associated chatbot called Claude that is roughly comparable to OpenAI’s ChatGPT or Google’s Bard.
San Francisco-based Anthropic, which has established itself as a public benefit corporation (PBC), has raised around $10 billion since its inception. Aside from the $4 billion from Amazon, Anthropic also counts Google as a big-name investor, with more than $2 billion from Alphabet’s subsidiary. The CMA has also launched an early-stage “invitation to comment” on Google’s investment that is still pending.
The CMA had been looking at whether key aspects of the Amazon and Anthropic partnership would result in “Amazon having material influence over Anthropic.” This is part of a growing trend in the AI realm, where critics argue that Big Tech is seeking to gain control over startups by adopting a new M&A approach that stops short of a full acquisition. This so-called “quasi-merger” might include hiring startup founders and talent, or making strategic investments.
However, the CMA said a “relevant merger situation” had not been created under the provisions of the Enterprise Act 2002, meaning it didn’t even get to a point where it could assess whether Amazon has attained “material influence” over Anthropic. This is because Anthropic’s U.K. turnover doesn’t meet the £70 million threshold to qualify for investigation, and the companies collectively don’t “account for a 25% or more” share of supply of the goods or services in question.
“As we’ve made clear, Anthropic is an independent company and our strategic partnerships and investor relationships do not diminish our corporate governance independence or our freedom to partner with others,” an Anthropic spokesperson said.
This probe was one of many similar investigations launched by the CMA of late. It recently cleared Microsoft’s Inflection acqui-hire, but concluded that the deal was tantamount to a merger. Microsoft also dodged antitrust scrutiny for buying a stake in Mistral AI.
Separately, the CMA has an ongoing case against Microsoft’s close ties with OpenAI — it launched a formal “invitation to comment” for stakeholders last year, but there’s been no progress to report since.
