Openai collects $ 40 billion in agreements with SoftBank that appreciates it to $ 300 billion | Open

Openi has collected $ 40 billion (£ 31 billion) through fundraising led by the Japanese group Softbank, in an agreement that enhances the Chatgpt developer at $ 300 billion (£ 233 billion).

Openai said that the financing round would allow the company to “further push the frontiers of the research to”. He added that the support of SoftBank “would open the way” towards agi, or artificial general intelligence, the term for artificial intelligence systems that can combine or overcome humans in almost all cognitive tasks.

“Hundreds of millions of people use the chatgpt every week,” said the CEO of Openai, Sam Altman. “This investment helps us to push the border and make the IA more useful in everyday life.”

The fundraiser came when Altman said that Chatgpt added a million users in just an hour while his new function of generation of images takes off with consumers, accompanied by a viral obsession for the recreation of art in the studio Ghibli.

SoftBank must put $ 10 billion at the beginning in Openni and $ 30 billion more by the end of 2025 if certain conditions are satisfied. The Japanese company will provide 75% of funding in the last round, with the rest provided by Microsoft -An key supporter of Openi-and the investment companies focused on Coatue Management, Altimeter Capital and Thrive Capital technology.

“Artificial intelligence is a distinctive force that model the future of humanity,” said Masayoshi Son, president and CEO of Softbank. “Our partnership expanded with Openi accelerates our shared vision to unlock its full potential.”

Openai also announced on Monday that he was building a more open artificial intelligence model as he faces growing competition in the open source space from Deep -week And destination, whose models are freely available and, unlike chatgpt, can be regulated. Openi is preparing to launch an “open weight” model that can be developed by users.

Openi had been a ferocious defender of closed models and owners who do not allow developers to change the basic technology to adapt artificial intelligence to their objectives. “We have been thinking about it for a long time, but other priorities have had precedence. Now it is important to do,” said Altman.

Openai and defenders of closed models-which include Google-have often denounced open models as more risky and more vulnerable to the use of the harmful and non-US government.

Elon Musk, a former Openi investor who Recently guided An offer for the consortium of almost $ 100 billion for the company, has Called Openi “Go back to the open strength and focused on safety for the good”.

Large companies and governments have proved reluctant to use artificial intelligence models on which they have no control, especially when data security is a concern. Meta and Deepseek allow companies to download and modify their models.

The CEO of Meta, Mark Zuckerberg, said this month that his large Ai Llama model hit 1 billion download. The low -cost R1 model of Deepseek has shaken the world of artificial intelligence in January.

Openi has ridden the success of his latest generation features of chatgpt images. In addition to marking the increase in user numbers, Altman said that the new image tool overwhelmed Openi graphic processing units.

Images that replicate art from the Japanese study Ghibli have proliferated from the launch of the new functionality. Altman has indicated that the function has fewer restrictions in use, with the tool that represents a “high -level brand for us in allowing creative freedom”.

A spokesman for Openii said: “We continue to prevent generations in the style of individual living artists, but we allow wider study styles – that people have used to generate and share some truly delicious and inspired original creations of original fans”.

With France-Press Agence