Every vendor briefing on AI eventually arrives at the same reassurance: we are here for the long term, our roadmap is funded, our enterprise commitment is not going away. It used to be easy to accept at face value. The numbers coming out of OpenAI this week make it harder.

The Information, citing people with direct knowledge, reports a genuine disagreement between Sam Altman and Chief Financial Officer Sarah Friar over when and whether to take OpenAI public. Altman has privately said he wants a stock market listing as early as the end of 2026. Friar has raised concerns. The reason those concerns matter: OpenAI is expected to spend more than $200 billion before it starts generating more cash than it burns, and Altman has already committed the company to $600 billion in spending over the next five years. Going public in that position is not a sign the company has arrived. It is a way to raise more money from a new class of investors — ordinary shareholders — to keep funding losses that private backers can no longer cover alone.

For any company buying AI services from OpenAI, that shift has practical consequences. Once a company lists on a stock exchange, its decisions get scrutinized every three months on an earnings call. Investors want to see either a clear path to profit or fast enough revenue growth to justify the losses. Both of those pressures tend to produce the same outcomes for enterprise customers: prices go up or stay firm, long-term contract terms tighten, and any product investment that does not produce near-term revenue gets deprioritized. None of that makes OpenAI a bad partner. It does mean the relationship changes in ways worth thinking through before you have built three years of internal tooling on top of their platform.

When the CFO pushes back, pay attention

The fact that Friar has raised concerns at all is the signal worth tracking. She is not someone who defaults to caution. As Chief Executive Officer of Nextdoor she took a company public under difficult market conditions. As Chief Financial Officer of Block she managed the finances of a company that was simultaneously running a payments business, a consumer lending operation, and a music streaming service. Someone with that background pushing back on IPO timing is doing the math on what the revenue curve actually looks like versus what investors will be willing to pay for.

Friar is not the only senior executive at OpenAI whose instincts matter here. Fidji Simo, who joined as Chief Executive Officer of Applications after leading product at Meta for years, brings a different kind of discipline: the experience of scaling consumer products into enterprise-grade platforms, understanding what it takes to build a business around AI that goes beyond research demos. The combination of Friar's financial rigor and Simo's operational experience is exactly what OpenAI needs to temper Altman's speed and ambition.

The closest parallel in recent technology history is Ruth Porat at Google. When she joined as Chief Financial Officer in 2015, coming from Morgan Stanley, Google was flush with search advertising revenue but spending freely on moonshots — self-driving cars, internet balloons, life extension research — with little public accountability for returns. Porat brought Wall Street discipline to a company run by engineers who had never had to answer hard questions about capital allocation. She cut spending on projects not meeting financial expectations and reined in the "Other Bets" division, and the stock responded. Her reputation became shorthand for "Wall Street grit meets Silicon Valley innovation." OpenAI needs that same dynamic. Altman is a visionary builder in the mold of Page and Brin. What the company needs, badly, is for Friar and Simo to be empowered to play the Porat role — not to slow the ambition, but to make it legible to the market and sustainable beyond a single person's force of will.

The $600 billion spending commitment is not a forecast. It is contracts and deals already signed — with chip manufacturers, data center builders, and infrastructure partners. $200B+ projected losses before profitability is not an accounting footnote. It describes the business model as it currently stands. Friar's job is to find the money to fund that on terms that do not hurt existing investors. Listing shares before the revenue story is strong enough to justify the valuation is one of the faster ways to get that math wrong.

The spending that sits underneath all of this

The $600 billion figure is not abstract. It connects directly to the Stargate project — a $500 billion data center build backed by SoftBank, Oracle, and others — plus multibillion-dollar agreements with Nvidia, AMD, Broadcom, and Cerebras to secure computing capacity across multiple years. OpenAI is not spending on office space and salaries. It is building physical infrastructure at a scale that resembles a national utility, not a software company. The revenue model that makes that worthwhile requires AI to become as routine and necessary to business operations as email, and to be priced to match.

There are signs that is happening. In the first quarter of 2026, the number of new apps submitted to Apple's App Store jumped 84% compared to the same period a year earlier, driven in significant part by tools like Claude Code and OpenAI's Codex that let people build software without deep technical training. Developer momentum is real. The question is whether it converts to the kind of predictable enterprise revenue that a public company needs to show investors.

Altman moves fast. That is a genuine strength and the reason OpenAI exists at the scale it does. Friar asking hard questions about timing is exactly the right job for a CFO. The fact that this disagreement has surfaced publicly, even through anonymous sources, suggests it has not resolved quietly. For organizations that have built or are planning to build significant internal capabilities on top of OpenAI's technology, that unresolved question about the company's financial strategy deserves attention.

ChatGPT is a consumer product. Enterprise buyers need something different.

Millions of people use ChatGPT and think of it as synonymous with AI itself. That level of brand recognition is rare and valuable. It is also a problem when the same brand needs to convince a hospital system's legal team, a bank's security committee, or a government agency's procurement office that it is a trustworthy long-term infrastructure partner.

Consumer products are built for speed, ease of use, and broad appeal. Enterprise technology is bought on the basis of security, predictability, compliance with regulations, and accountability when things go wrong. ChatGPT's public identity, shaped by years of consumer marketing and high-profile controversies including the boardroom crisis of 2023, is misaligned with what enterprise buyers need to feel confident. OpenAI has been trying to serve both audiences with the same story, and the gap is showing.

Anthropic has built its entire identity around the enterprise case. Safety, governance, and secure deployment have been the company's public position since it was founded. To close the gap, OpenAI needs to demonstrate two things Anthropic has already operationalized: that its technology can be controlled and audited by the customer, and that it can run inside a customer's own secure environment rather than only through OpenAI's cloud.

Sam Altman is a great communicator. That cannot be the whole marketing plan.

Altman is one of the most effective technology executives at shaping public narrative. His appearances, interviews, and social media presence have driven enormous awareness of OpenAI and, by extension, of AI as a category. The company has benefited enormously from having a founder who can explain the technology compellingly to a general audience.

The risk is overexposure and over-reliance. When one person is the primary face of a company's messaging, every controversy that involves them becomes a brand event. The late 2023 board crisis damaged OpenAI's reputation not just because of what happened internally, but because the company's public identity was so bound to Altman personally that his five-day absence destabilized everything customers and partners could see from the outside.

Anthropic has taken a different approach to building enterprise credibility — not through a single spokesperson but through structured partnerships. The company has committed $100 million to a partner network in 2026 and trained 30,000 professionals at Accenture alone on how to deploy Claude inside client organizations. That is how enterprise technology builds lasting market presence: through the people who implement it, not through the person who founded it.

OpenAI needs a dedicated enterprise marketing team with its own voice, its own customer references, and its own roadmap narrative that stands apart from the consumer ChatGPT story. Institutional buyers, procurement teams, and CIOs evaluating a multimillion-dollar commitment need to hear from people who understand their problems specifically — not from a general-audience interview about the future of artificial general intelligence.