Some early investors in OpenAI are questioning the start-up’s $852 billion (£628bn) valuation, amid shifts in strategy that make it appear unfocused, the Financial Times reported.

Some investors are unhappy with OpenAI’s shift in strategy to target higher-margin enterprise sales, an area in which it trails behind competitor Anthropic, saying this could leave OpenAI vulnerable to Anthropic and Google.

“You have ChatGPT, a 1 billion-user business growing 50-100 per cent a year, what are you doing talking about enterprise and code? It’s a deeply unfocused company,” an unnamed early backer of OpenAI told the paper.

An investor who has backed both OpenAI and Anthropic said that an investment into OpenAI’s most recent funding round would have to assume an IPO valuation of $1.2tn or more, which has become more difficult to justify given the cheaper proposition of buying into Anthropic, valued at $380bn.

OpenAI’s purchase of tech talk show TBPN was criticised by an OpenAI investor as “a distraction”.

The company’s recent shifts have included shuttering its video generation tool Sora, which eliminated a $1bn investment from Disney, scrapping plans for an “adult” chatbot, drastically paring back an investment deal with Nvidia, and halting plans to develop a $30bn data centre in the UK and extend a site in Abilene, Texas.

Meanwhile the company has switched to pushing its Codex coding tool to businesses, in direct competition with Anthropic.

Jai Das, president of investment firm Sapphire Ventures, who is not an investor in OpenAI or Anthropic, referred to OpenAI as potentially “the Netscape of AI”, referring to the browser darling of the late 1990s that was superseded by Microsoft and was eventually bought by AOL.

Others pointed out that OpenAI holds a strong lead over Anthropic in procuring computing resources, while the company’s chief financial officer Sarah Friar said its large recent funding round shows the confidence of investors.