OpenAI is worth more on paper than almost any company in history, but a growing faction of its own investors isn't convinced the number reflects reality.
The world's most talked-about AI company is sitting on an $852 billion valuation and a mounting credibility problem. Following a secondary share sale earlier this year that briefly made OpenAI the most valuable private company on the planet, some of its own backers are quietly questioning whether that figure can hold. The concern isn't about the technology. It's about whether OpenAI can convert its brand dominance into the kind of stable, recurring revenue that justifies a number more commonly associated with publicly traded industrial giants.
The pressure is pushing the company into a full pivot toward enterprise customers, a strategic shift that CEO Sam Altman is personally championing. Altman has made headlines in recent months declaring that artificial general intelligence is within reach, but behind the scenes the operational focus is considerably more mundane: signing large-scale corporate contracts, expanding security and compliance offerings, and building the B2B infrastructure that sustains a real business. It's a significant departure from the consumer-first identity that made ChatGPT a cultural phenomenon, and not every investor signed up for that version of OpenAI.
The strategic urgency is amplified by Anthropic, which has spent the past year methodically picking off enterprise accounts that might otherwise have defaulted to OpenAI. Founded by Dario and Daniela Amodei, both former OpenAI executives, Anthropic has positioned its Claude models as the responsible choice for corporate procurement teams navigating governance and compliance requirements. That pitch is landing. Legal departments, financial institutions, and regulated industries are increasingly running Claude evaluations alongside or instead of GPT-4 class models, and Anthropic's revenue trajectory reflects it.
The sibling rivalry framing understates what's actually happening competitively. Anthropic isn't simply offering a safer alternative to OpenAI's products. It's building an enterprise sales motion, partner ecosystem, and regulatory credibility story designed to make Claude the institutional default in sectors where reliability and auditability matter more than raw capability benchmarks. OpenAI's response has been to accelerate its own enterprise offerings, but catching up in trust and process is slower work than shipping a new model.
Burn Rates and the Path to Profitability
Investor skepticism sharpens further when the cost structure enters the conversation. Training frontier models at OpenAI's scale is extraordinarily expensive, and the company has consistently operated at high burn rates while the AI sector has shifted from rewarding ambition to demanding unit economics. Backers who entered at lower valuations may still be sitting on paper gains, but those who participated in more recent rounds are doing math that requires OpenAI to not only grow but to grow faster than its costs. That calculation is harder when a well-funded competitor is eroding the enterprise market share the growth story depends on.
Microsoft's partnership remains a structural asset, providing distribution and cloud infrastructure that Anthropic cannot easily replicate at scale. But partnerships don't automatically translate into customer retention, and enterprise clients are sophisticated enough to run parallel evaluations regardless of what their cloud vendor prefers. The Microsoft relationship is a floor, not a ceiling.
What to watch in the coming quarters is whether OpenAI's enterprise conversion rates actually move. Valuation at this level demands not just revenue growth but a demonstrable moat: customers who renew, expand, and don't defect to Claude the moment a procurement cycle opens up. If Anthropic continues compounding its enterprise credibility while OpenAI manages the complexity of restructuring a company that was never built around B2B sales, the $852 billion figure will face its most serious test yet from the funding market itself.
Also read: Meta locks in Broadcom through 2029 to build the custom chips that will power its AI ambitions across Facebook, Instagram and WhatsApp • X rolls out live Cashtags to connect stock and crypto conversations with real-time market data • YouTube pulled an Iran-linked AI Lego channel with nearly 3 million views after Microsoft traced it to the Revolutionary Guard
