OpenAI hires new CRO as executive shake-up continues

OpenAI just fired its chief revenue officer after nine months. That sentence alone should alarm anyone watching the company that claims to be building the future of intelligence. Denise Dresser didn't fail because she couldn't sell. She failed because the product still doesn't sell itself the way Sam Altman promised it would.

Enter Dali Rajic. His résumé: president and COO of Wiz, a cloud security company Google swallowed for $32 billion — the search giant's largest acquisition ever. That is not the background of a man hired to evangelize AGI. That is the background of a man hired to close seven-figure deals with CISOs and procurement committees. The message is unmistakable: OpenAI is done pretending it's a research lab. It wants to be a sales machine.

The timing is deliberate. In the past month, the company has lost its COO, Brad Lightcap, and its second-in-command, Fidji Simo, who held the title CEO of AGI Deployment — a role that now sounds like a contradiction in terms. Greg Brockman, a co-founder long overshadowed by Altman, has stepped into the management vacuum. He announced Rajic's hire in a blog post that praised Dresser's "formative period" leadership, then immediately undercut her by saying the deployment model is "changing rapidly" and Rajic will deliver "repeatable execution." Translation: we have no repeatable execution.

One billion weekly active users. Two million businesses. Those are the numbers OpenAI touts. They are staggering. They are also, by the company's own admission, not translating into the revenue trajectory insiders expected. Executives have said as much, privately and publicly. The growth is real. The monetization is not. When a company with a billion users cycles revenue chiefs like rental cars, the problem isn't the product. The problem is the business model.

The IPO filing is confidential. The $7 billion tender offer for employee shares closed this week. Both moves usually precede a public listing. But the tender offer — letting employees cash out before an IPO — often signals the IPO is further out than leadership wants to admit. You don't build a $7 billion liquidity bridge for a sprint. You build it for a marathon. The market window is tightening. Interest rates are sticky. AI fatigue is setting in among institutional investors. OpenAI needs a cleaner story than "we have a billion users and we're figuring out the rest."

Altman has spent this year loudly pivoting to enterprise. He has cut research projects he deemed distractions. He has reorganized around deployment. Yet the revenue chief lasts nine months. The COO leaves. The AGI deployment chief leaves. The president steps in. The new sales boss comes from a company whose exit was a security acquisition, not an AI platform. None of this resembles a company executing a clear plan. It resembles a company reacting to pressure.

The blog post originally included a phrase about "relentless focus on measurable business impact." That phrase disappeared from the published version. Someone — likely legal, possibly Altman — decided the words invited scrutiny the numbers couldn't support. That edit is more honest than the press release. It admits the focus isn't relentless. It isn't measurable. The impact is still theoretical.

Rajic knows how to sell to the Fortune 500. He knows how to build a sales org that survives quarterly reviews. He does not know how to sell a product that hallucinates, leaks data, and lacks deterministic guarantees — the very problems that make enterprise buyers hesitate. His Wiz experience solved a different problem: securing cloud infrastructure. OpenAI's problem is convincing enterprises that probabilistic output is worth deterministic budget.

The shake-up will continue. You don't stabilize a revenue organization by replacing its leader three times in a year. You stabilize it by delivering a product that salespeople can quote, implement, and renew without engineering support tickets. OpenAI doesn't have that product yet. Rajic cannot conjure it. Brockman cannot blog it into existence. Altman cannot promise it into the next funding round.

The billion users are a trap. They create the illusion of product-market fit. They mask the absence of product-revenue fit. Every executive departure since Dresser arrived confirms the mask is slipping. The next CRO will inherit the same structural problem: a miracle model that enterprises still treat as a pilot project. Until that changes, the revolving door stays in motion.