Microsoft and Nvidia to invest up to $19.5 billion in Anthropic amid AI bubble fears
Sign up now: Get ST's newsletters delivered to your inbox
Investors have grown increasingly concerned about these so-called circular AI deals, which trigger bubble fears.
PHOTO: REUTERS
Follow topic:
New York - Microsoft and Nvidia are committing to invest up to a combined US$15 billion (S$19.5 billion) in Anthropic, in a move that ties the artificial intelligence (AI) developer closer to two of the biggest backers for its rival OpenAI.
The investment will be part of Anthropic’s next funding round, according to a person familiar with the matter. Anthropic has also committed to purchase US$30 billion of computing capacity from Microsoft’s Azure cloud service, the companies said on Nov 18.
The deal is part of a growing number of tie-ups that have seen cloud computing and chips providers back leading AI developers, which in turn spend money on services from those same firms. Investors have grown increasingly concerned about these so-called circular AI deals, with some worrying they are a clear sign of a bubble.
Microsoft shares slid about 3 per cent and Nvidia fell about 1 per cent on Nov 18 amid a broader market pullback.
For Anthropic, the arrangement also brings it closer to Microsoft, which has invested more than US$13 billion in OpenAI and played a pivotal role in the ChatGPT maker’s success. In recent months, however, Microsoft and OpenAI have been in more open competition with one another.
“We are increasingly going to be customers of each other – we will use Anthropic models, they will use our infrastructure, and we will go to market together,” said Microsoft chief executive Satya Nadella of Anthropic in a video posted on Nov 18. “Of course, this all builds on the partnership we have with OpenAI, which remains a critical partner for Microsoft.”
For Nvidia, the deal binds its technology to another one of the companies leading the race to make AI part of everyday life. Nvidia is throwing more money at end users of its chips at a time when investors are increasingly concerned that AI is not yet generating enough revenue or profits to justify the massive spending on new infrastructure. The company is set to report earnings results on Nov 19.
Nvidia, the world’s most valuable publicly traded firm, will invest up to US$10 billion in Anthropic, the chipmaker said. Anthropic, meanwhile, is initially committed to taking as much as one gigawatt of computing capacity that will be based on Nvidia’s Grace Blackwell and Vera Rubin chip designs. Nvidia’s engineers will work with their counterparts at Anthropic to help make future chip designs work better with the start-up’s AI software and services.
Anthropic, founded in 2021 by former employees of OpenAI, has positioned itself as a reliable, safety-conscious firm that users can trust. While it is smaller than OpenAI, the company’s Claude chatbot and underlying technology have gained traction with enterprise customers in sectors like finance and healthcare, as well as with developers.
The company recently raised US$13 billion at a US$183 billion valuation in September and says it has 300,000 business customers.
Like rival OpenAI, Anthropic is now accelerating its push to build out more infrastructure to support AI development and broader adoption of the technology. Anthropic plans to spend US$50 billion to build custom data centres for AI work in several US locations, including Texas and New York.
Anthropic also announced a deal in October with Alphabet’s Google
Amazon is also a key Anthropic partner. The Seattle-based computing giant has backed the AI start-up with an investment of US$8 billion and built a massive complex of data centres and custom Amazon Web Services AI chips for Anthropic’s use. Amazon remains Anthropic’s primary cloud provider. BLOOMBERG
