The plan, which has not been finalised, involves Google providing cloud computing services to Anthropic, according to the people, who asked not to be named because the information is private.

The deal will allow Anthropic to use Google’s tensor processing units, or TPUs – the company’s chips that are custom designed to accelerate machine learning workloads, one of the people said. Google is a previous investor in, and cloud provider for, Anthropic.

Anthropic and Google declined to comment. Talks are in their early stages and the details could change. Google shares jumped more than 3.5% in extended trading, while Amazon.com Inc, also an Anthropic investor and cloud provider, slipped about 2%.

Founded in 2021 by former OpenAI employees, San Francisco-based Anthropic is best known for its Claude family of large language models, which compete with OpenAI’s GPT models.

Like its peers, Anthropic has been raising significant sums to keep pace in a race to advance AI, which industry leaders say will require more resources for research and breakthroughs, as well as consumer demand.

Anthropic recently held early funding talks with Abu Dhabi-based investment firm MGX, about a month after closing a US$13 billion funding round.

That financing, led by Iconiq Capital, with Fidelity Management and Research Co and Lightspeed Venture Partners as co-leads, nearly tripled Anthropic’s valuation to US$183 billion, including dollars raised.

Google previously invested about US$3 billion in Anthropic; the tech giant committed to invest US$2 billion in the AI startup in 2023, and followed up with another US$1 billion investment early this year.

Amazon has committed to invest about US$8 billion in Anthropic, which is a key AI customer of Amazon Web Services, and a major user of Amazon’s custom AI chips.