Google investing up to $40 billion more in OpenAI rival Anthropic

US tech company Google is investing $10 billion in Anthropic, the OpenAI rival said on Friday.

A further $30 billion may follow if the development firm behind the chatbot Claude meets agreed targets, it said.

Google has for years been one of Anthropic’s investors and supplies the company with specialist chips for artificial intelligence (AI).

The step comes days after Amazon invested $5 billion in Anthropic – with the prospect of a further $20 billion in later stages.

The cash injection is an example of circular deals in the AI sector, in which investments flow back to the investor, at least in part, in return for chips or computing power.

Anthropic – and ChatGPT developer OpenAI – urgently need more computing capacity for their AI software. Both companies are also preparing to go public.

Anthropic has repeatedly made headlines in recent months and risked conflict with the US government as, despite massive pressure, it insisted its AI models must not be used in autonomous weapon systems or for mass surveillance in the United States.

The Pentagon then declared Anthropic a supply chain risk, which could severely hamper the use of the company’s software by government agencies. The company is taking legal action in response.

Anthropic recently demonstrated the importance of its technology to the US government. A new Anthropic model called Claude Mythos Preview is identifying vulnerabilities in a series of software programmes, some of which have remained undiscovered for decades.

In the wrong hands, it would be a dangerous cyber weapon. It could give the US government an advantage in protecting its own programmes while also allowing it to exploit potential vulnerabilities for espionage abroad.

Anthropic has no plans to make Mythos Preview public.