Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
Amazon has committed a further $2.75bn to artificial intelligence start-up Anthropic, strengthening its ties to the company as Big Tech’s race to dominate the booming AI sector intensifies.
The deal is Amazon’s largest ever venture investment and brings its total commitment to Anthropic to $4bn. The partnership is one of a number struck between AI start-ups and technology giants over the past year, as Google, Microsoft, Amazon and others bet that strength in generative AI will be their biggest competitive edge for years to come.
Amazon invested an initial $1.25bn in September last year, and had reserved the right to increase that total to $4bn before the end of March.
Anthropic is separately raising hundreds of millions of dollars from venture capitalists. Combined, the deals could ultimately furnish Anthropic with billions of dollars in capital and credits to use cloud and computing services, at a valuation of more than $18bn, according to people with knowledge of the discussions.
Amazon is increasing its bet on Anthropic even as scrutiny from regulators in the US, UK and EU has grown over concerns that deals between Big Tech and the most prominent AI start-ups may pose anti-competitive risks.
Microsoft has taken an early lead in the generative AI space, committing $13bn to OpenAI, whose ChatGPT chatbot ignited a frenzy of investor excitement about the fast-developing technology. Microsoft has since invested in European AI start-up Mistral and last week struck a $650mn deal to hire the leadership team and researchers from AI company Inflection, according to a person with knowledge of the deal.
Amazon’s Anthropic deal gives the Seattle-based company a well-known partner in AI as it vies with rivals Microsoft and Google for dominance in the space.
Amazon reiterated on Wednesday that it would be Anthropic’s “primary cloud provider” for key workloads, which will run on Amazon Web Services. Anthropic will also use AI computer chips developed in-house by Amazon, and is working with the cloud provider to develop and improve the hardware, which aims to compete with Nvidia’s highly sought-after chips.
However, the deal does not tie Anthropic to Amazon’s cloud exclusively, as OpenAI’s partnership with Microsoft does, and the start-up will continue to also use Google’s cloud. Google has agreed to invest up to $2bn into Anthropic, according to a person with knowledge of that deal.
Anthropic was founded in 2021 by a group of researchers who quit OpenAI over concerns about its commitment to AI safety. The San Francisco-headquartered company is widely regarded as the most likely challenger to OpenAI’s early dominance of the sector. Its latest model, Claude 3, outperforms OpenAI’s GPT4 on a number of industry benchmarks.
The company is selling shares at $30 apiece in an effort to raise cash and computing credits to fund the costly development of cutting-edge AI models, according to people with knowledge of the deal.
Venture capital investors are also expecting to commit at least $750mn to the company in a round that is expected to close in April, according to the people. The bulk of that total has been raised via a special purpose vehicle set up by Silicon Valley firm Menlo Ventures, an existing Anthropic investor, they added.
One of Anthropic’s early investors was Sam Bankman-Fried’s failed cryptocurrency group FTX, which has since sold the bulk of its stake in the company as part of bankruptcy proceedings. A fund owned by Abu Dhabi’s sovereign wealth fund Mubadala agreed to buy the majority of FTX’s stake for $500mn, according to court filings last week.