Amazon.com Inc said on Monday it would invest up to $4 billion in cash in high-profile startup Anthropic to compete with growing cloud computing rivals in artificial intelligence. As part of the deal, Amazon employees and cloud computing customers will have early access to Anthropic's technology, which they can inject into their own businesses.
San Francisco-based Anthropic has also pledged to rely heavily on Amazon's cloud services, including training its future artificial intelligence models on a raft of proprietary chips purchased from Amazon.
In a joint interview, the CEOs of Amazon's cloud computing unit AWS and Anthropic said Amazon will immediately invest $1.25 billion in Anthropic, and either party has the right to trigger another $2.75 billion in capital injections from Amazon.
They declined to say how much of Anthropic's stake Amazon will now own, or the startup's latest valuation (its last funding round valued it at more than $4 billion). Amazon said it will not get a seat on Anthropic's board of directors and its stake amounts to a minority stake.
The news may be Amazon's biggest response yet to challenges from Microsoft and Google. Both smaller cloud computing rivals have launched or developed powerful artificial intelligence this year. The deal also signals Amazon's cloud computing unit is working to connect with artificial intelligence startups that are reshaping its industry.
For Amazon, Monday's deal could mean rising demand, including for chips that power artificial intelligence. Anthropic, for example, agreed to develop technology for Amazon's in-house Trainium and Interentia chips.
Adam Selipsky, CEO of Amazon's AWS division, said the agreement "will help improve Anthropic's model and help improve our chip technology and artificial intelligence infrastructure."
Anthropic CEO Dario Amodei said his company "secured this funding in a way that prioritizes security" and "allows us to continue scaling our model."