AI: OpenAI Pre-Markets its IPO Through the Churn. AI-RTZ #1182

Yesterday we walked through Anthropic’s weekend of hard numbers on the road to a fall mega-AI IPO. Its annualized revenue is topping $65 billion before its highly anticipated mega-AI IPO later this year.

Today, the other side of the parade in this AI Tech Wave: OpenAI spent the same week pre-marketing its own listing, with an investor meeting full of milestones, and a C-suite churning at a pace that made its own headlines.

Not dissimilar to the senior people shuffling at OpenAI’s other arch-competitor Google, of late.

Start with the business, because the business is genuinely moving. CFO Sarah Friar held a Friday meeting with current shareholders, and CNBC lays it out in ” OpenAI CFO Friar tells investors that enterprise business now bigger than consumer by revenue”:

“We entered the year at 60-40, but enterprise has accelerated much faster than expected and those lines have now crossed… The majority of our revenue is now enterprise.”

That enterprise crossover arrived ahead of the company’s own end-of-2026 parity forecast. The annualized run rate has hit $40 billion, up 20% month over month in July, with business customers growing faster still at 32%. Advertising, which began testing in ChatGPT in February, is approaching a $1 billion run rate of its own. And per the FT, much of the recent growth has come since GPT-5.6 shipped five weeks ago.

Then the line that caught our eye in particular. Friar told the room that enterprise customers have moved on from racking up unmetered AI bills:

“Enterprise customers have moved from tokenmaxxing to focusing on cost per unit of intelligence.”

Regular readers know ‘Tokenmaxxing’ from these pages: we wrote about customers starting to ask ‘how much’ for AI tokenmaxxing in May, and Meta stepping back from it in June. When the CFO of the a major consumer AI company frames her IPO pitch around the end of that era, and around ‘cost per unit of intelligence’, that is the y-axis of our supply and demand series showing up in an investor deck. Cheaper intelligence, measured per unit of work, with demand expanding to meet it.

Now the senior exec churn. The FT counts nearly half a dozen reorganizations this year: chief revenue officer Denise Dresser out after eight months, longtime executive Brad Lightcap out after eight years, ethics chief Chloé Bakalar out inside a year, Fidji Simo stepped back to an advisory role for medical reasons. And the ‘preparedness’ safety team disbanded into existing groups.

The FT also reports the IPO, once expected this year, is now likely next year, at a valuation that could reach $1 trillion, and that staff have been cashing out into a near-$7 billion tender at the current $852 billion mark.

The two readings of that churn both deserve discussion.

The internal one, per FT sources: frustration, burnout, and ‘politics’ in the run-up to one of the biggest listings ever. The investor one, from a large OpenAI backer:

“I don’t think it’s uncommon to clean up the organisation before going public. You want to know who your people are.”

Both can be true at once. We discussed the Dresser-to-Dali-Rajic handoff on Friday’s ARD, including the Wiz pedigree Rajic brings to the enterprise fight.

My take: both mega-AI IPOs are moving aggressively, and that is worth noting.

Anthropic is pre-marketing a 2028 revenue projection; OpenAI is pre-marketing an enterprise crossover, an ads line, and model efficiency, its own version of the margin math. Note who set the terms of that fight: per the FT, Altman’s directive to staff is to cut ‘side quests’ and compete with Anthropic for business clients. The company that was ahead on revenue at $24 billion entering the year now trails its rival’s $65 billion-plus run rate, and trails it to the IPO window too. We assessed OpenAI’s IPO prospects six months early back in March, and flagged that tight clock in June. It is now the operating reality.

The push and pull to watch from here: consumer scale, a billion-plus weekly users with a small paying share plus a growing ads line, is OpenAI’s distinct asset, and the hard to value.

The senior exec churn is the cost of professionalizing a research lab into a public company on a deadline. And massive global media focus.

And the disbanding of the preparedness team, whatever the reorganization logic, hands the bears a governance talking point at exactly the moment the roadshow needs less of those issues.

The overall read cuts both ways, and worth continued focus.

A key part of the AI Tech Wave worth tracking. Stay tuned.

(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here.)