OpenAI has raised $40bn (£31bn) through fundraising led by the Japanese group SoftBank, in a deal that values the ChatGPT developer at $300bn (£233bn).

OpenAI said the funding round would allow the company to “push the frontiers of AI research even further”. It added that SoftBank’s support would “pave the way” towards AGI, or artificial general intelligence, the term for AI systems that can match or exceed humans at nearly all cognitive tasks.

“Hundreds of millions people use ChatGPT each week,” said OpenAI’s chief executive, Sam Altman. “This investment helps us push the frontier and make AI more useful in everyday life.”

The fundraising came as Altman said ChatGPT had added a million users in just one hour as its new image generation feature takes off with consumers – accompanied by a viral obsession with recreating Studio Ghibli-style art.

SoftBank is to put $10bn at first into OpenAI and $30bn more by the end of 2025 if certain conditions are met. The Japanese company will provide 75% of the funding in the latest round, with the rest provided by Microsoft – a key backer of OpenAI – and the tech-focused investment firms Coatue Management, Altimeter Capital and Thrive Capital.

“AI is a defining force shaping humanity’s future,” said Masayoshi Son, the chair and chief executive of SoftBank. “Our expanded partnership with OpenAI accelerates our shared vision to unlock its full potential.”

OpenAI also announced on Monday that it was building a more open generative AI model as it faces growing competition in the open-source space from DeepSeek and Meta, whose models are freely available and, unlike ChatGPT, can be adjusted. OpenAI is preparing to launch an “open weight” model which can be fine-tuned by users.

OpenAI had been a fierce defender of closed, proprietary models that do not allow developers to modify the basic technology to adapt AI to their goals. “We’ve been thinking about this for a long time, but other priorities took precedence. Now it feels important to do,” said Altman.

OpenAI and defenders of closed models – which include Google – have often decried open models as riskier and more vulnerable to malicious and non-US government use.

Elon Musk, a former OpenAI investor who recently led a near-$100bn consortium bid for the company, has called on OpenAI to “return to the open-source, safety-focused force for good it once was”.

Large companies and governments have proven reluctant to use AI models they have no control over, especially when data security is a concern. Meta and DeepSeek let companies download and modify their models.

Meta’s chief executive, Mark Zuckerberg, said this month that its large language AI model Llama had hit 1bn downloads. DeepSeek’s lower-cost R1 model rocked the world of artificial intelligence in January.

OpenAI has been riding on the success of its latest image-generation features in ChatGPT. As well as flagging the boost in user numbers, Altman has claimed that the new image tool has overwhelmed OpenAI’s graphics processing units.

Images replicating art from Japan’s Studio Ghibli have proliferated since the launch of the new feature. Altman has indicated that the feature has fewer restrictions on use, with the tool representing a “high-water mark for us in allowing creative freedom”.

An OpenAI spokesperson said: “We continue to prevent generations in the style of individual living artists, but we do permit broader studio styles – which people have used to generate and share some truly delightful and inspired original fan creations.”

With Agence France-Presse