OpenAI has announced that it has secured $40 billion in a funding round that values the artificial intelligence powerhouse at an unprecedented $300 billion, making it the largest capital raise ever for a startup. The investment, spearheaded by Japan’s SoftBank Group, will fuel OpenAI’s continued advancements in artificial intelligence research and development.
In a statement on its website, the San Francisco-based company said the partnership with SoftBank would allow it to “push the frontiers of AI research even further” and help develop AI systems that enhance scientific discovery, personalised education, and human creativity. The ultimate goal, according to OpenAI, is achieving artificial general intelligence (AGI), a system with human-level intelligence that benefits all of humanity.
SoftBank echoed this sentiment in its statement, saying it aims to realise Artificial Super Intelligence (ASI), which would surpass human intelligence. The Japanese investment firm described OpenAI as the closest company to achieving AGI and noted that significant computational power is required to reach this milestone.
As part of the investment, SoftBank will inject an initial $10 billion into OpenAI, with an additional $30 billion expected to follow later this year, subject to certain conditions. OpenAI plans to use the funding to expand its infrastructure and continue delivering advanced AI tools to the 500 million people who use ChatGPT weekly.
The funding announcement coincided with OpenAI’s revelation that it is working on a more open generative AI model, a move that marks a strategic shift for the company. Previously, OpenAI has staunchly defended its closed, proprietary AI systems, arguing that open-source models pose greater risks, particularly in the hands of malicious actors or foreign governments.
However, growing competition from rivals such as Meta and Chinese AI firm DeepSeek has put pressure on OpenAI to reconsider its stance. Many large corporations and governments have hesitated to adopt AI solutions built on models they cannot modify or control, especially due to data security concerns.
Unlike OpenAI’s closed approach, Meta’s Llama models and DeepSeek’s AI allow users to download and modify the underlying technology, granting them greater autonomy over data and customisation. Meta CEO Mark Zuckerberg recently announced that Llama has surpassed one billion downloads, while DeepSeek’s cost-effective R1 model, released in January, has shaken up the AI industry.
Meanwhile, OpenAI continues to experience soaring demand for its latest AI features, particularly its image-generation tools integrated into ChatGPT. Altman claimed on Monday that the new tool attracted one million users within an hour of its release. This surge in usage comes just days after he revealed that OpenAI’s graphics processing units (GPUs) were struggling to handle the overwhelming demand.
