OpenAI’s latest funding round, worth $8.3 billion, was oversubscribed five times. The investor appetite reflected their confidence in the AI startup’s ability to dominate a market that the UN Trade and Development projects will explode by 25 times in size in a decade.
OpenAI’s momentum is undeniable. The company has continuously upgraded its flagship ChatGPT product, recently launching GPT-5, which it claims can provide PhD-level expertise. Financially, its revenues have doubled in seven months, reaching $1 billion a month, with projections to hit $20 billion in annualised revenue by the end of the year.
The capital influx will primarily help OpenAI scale its compute infrastructure, particularly Stargate, a joint venture with Japanese investment firm SoftBank and technology company Oracle to build the world’s largest AI supercomputing infrastructure. OpenAI is also setting up its first data centre in Europe next year, which will house 100,000 Nvidia processors.
This infrastructure investment is critical as companies race to control the data centres and AI chips essential for training and operating advanced artificial intelligence models.
The numbers reflect this reality. Global data centre capacity surged from 20GW in 2016 to 57GW in 2024, with Goldman Sachs projecting 122GW by 2030. While OpenAI’s valuation reflects investor confidence, the fundraising itself underscores the infrastructure investments needed to maintain leadership in the AI market.
Challenger pack
OpenAI faces growing competition from well-funded AI startups. Anthropic, founded by former OpenAI employees, is nearing a $5 billion funding round that would value it at $170 billion, up from $61.5 billion in March. Elon Musk’s xAI has raised $10 billion at an $80 billion valuation and is seeking additional funding at a potential $200 billion valuation.
Venture capital funding to AI companies has exceeded $40 billion in each of the past three quarters, according to Crunchbase.
This financial backing is translating into competitive model performance. On the GPQA Diamond benchmark, which tests PhD-level science questions, xAI’s Grok 4 Heavy scored 88.9% and Anthropic’s Claude Opus 4.1 scored 80.9%.
The landscape shifted when Chinese startup DeepSeek released powerful open-weight models available for free. OpenAI released its own open-weight models in response. The competition now spans both proprietary and open-source approaches.
Incumbent advantage
OpenAI also faces pressure from the Big Tech firms. Meta, Google, Amazon, and Microsoft have collectively spent $291 billion over the past year, largely for AI infrastructure.
Last month, in a $2.4 billion deal, Google hired key executives from Windsurf, an AI coding company that OpenAI wanted to acquire. Google has also integrated ‘AI Overviews’ with its search engine, turning it into an “answer engine” that directly competes with the core function of chatbots like ChatGPT. This strategy leverages Google’s 2 billion monthly users and its market dominance.
Meta, meanwhile, is restructuring its AI division into Meta Superintelligence Labs. It has also acquired top-tier AI researchers from OpenAI, with multi-million-dollar compensation packages.
Partner paradox
OpenAI’s relationship with Microsoft, however, has turned complicated.
Microsoft, OpenAI’s primary backer with a $13.75 billion investment, is also a direct competitor seeking to lead the AI revolution. Copilot, Microsoft’s AI platform, boasts over 100 million monthly users. Microsoft’s server products and cloud services revenue jumped 27% year-over-year in the three months ended 30 June, driven by growth in Azure, its cloud or remote computing platform.
Microsoft holds crucial leverage as OpenAI attempts to convert into a for-profit company—a prerequisite for unlocking SoftBank funding and IPO plans. However, Microsoft has been withholding approval as both companies negotiate revising their contract, set to expire in 2030.
A major sticking point is a clause that could terminate Microsoft’s access to future OpenAI technology if the startup’s board declares that artificial general intelligence—AI’s capacity to learn and understand like humans and apply that knowledge to execute tasks—has been achieved.
This friction has real consequences: OpenAI’s attempt to acquire AI coding startup Windsurf failed because Microsoft’s IP rights would have extended to the new technology, which Windsurf rejected.
OpenAI needs capital to overcome these structural challenges and funding obstacles.
www.howindialives.com is a database and search engine for public data