(Bloomberg) -- Artificial intelligence startup Anthropic is in early talks to raise $2 billion, shortly after Amazon.com Inc. agreed to invest up to $4 billion in the company, people familiar with the matter said. 

The deal, which has not yet been finalized, represents an industrywide rush to claim a piece of Silicon Valley’s AI boom. 

The valuation for Anthropic has not yet been set, said the people, who asked to remain anonymous discussing private conversations. Tech news site the Information earlier reported on the company’s fundraising efforts, and said that a deal could value Anthropic at as much as $30 billion.  

If the funding round closes, it would be the latest huge financing for a startup building so-called foundation models, which take in large amounts of data to provide the basis of many AI tools. The models are extremely expensive to operate, meaning startups like Anthropic and its competitor OpenAI need ample resources to keep the technology running.  

The talks also represent a scramble by large tech companies to get a stake in the future of artificial intelligence. Earlier this year, Alphabet Inc.’s Google participated in a $450 million funding round for Anthropic. And in September, Amazon said that in addition to funding the startup, it would be Anthropic’s primary cloud computing provider and would supply chips. Meanwhile Microsoft Corp. has agreed to put more than $10 billion into OpenAI, a rival to Anthropic in generative AI.

Read More: Amazon to Invest Up to $4 Billion in AI Startup Anthropic

Anthropic’s corporate investors are eager to continue backing the startup, one person familiar with the talks said, but it’s possible the deal won’t close until a financial investor, such as a Wall Street firm or large venture capital investor, agrees to participate. That investor could be called on to determine the company’s valuation and lead the round. 

Founded in 2021 by former employees of OpenAI, Anthropic has positioned itself as being especially focused on responsible and safe technology. Its chatbot, Claude, is meant to be a kinder and gentler kind of AI, and can perform written tasks like summarizing, searching, answering questions and coding.

(Updates with context starting in the third paragraph.)

©2023 Bloomberg L.P.