As the OpenAI Turns

Somehow the company building artificial general intelligence has also become Silicon Valley's longest-running soap opera.

I started looking into the latest round of executive departures at OpenAI because I had a very simple question: is Sam Altman in trouble again?

Which is already a ridiculous sentence. Again? How many technology companies require that qualifier?

But that's OpenAI. Every time I think we're finally going to be allowed to just talk about the actual artificial intelligence, somebody leaves, somebody comes back, somebody restructures something, Microsoft gets dragged into it, the board changes, the company changes its corporate structure or Greg Brockman suddenly appears holding another set of keys.

At some point you have to admit it: OpenAI isn't just one of the most important companies in AI. It's the soap opera of AI. And holy crap, the writers have had a decade.

Previously, on OpenAI

OpenAI begins in 2015 as a nonprofit research organization built around a fairly extraordinary idea: artificial general intelligence could eventually become enormously powerful, so perhaps the organization building it shouldn't simply be another conventional corporation maximizing shareholder returns. Reasonable. Ambitious. Maybe even noble.

There was just one tiny problem. AGI is apparently expensive as hell.

By 2019, OpenAI had realized that donations were probably not going to pay for the mountains of compute required to pursue increasingly capable models.

So it invented something called a capped-profit company. Not quite nonprofit. Not quite normal for-profit. Investors and employees could make money, but returns were capped, while control ultimately remained with the nonprofit. If you're already squinting at the corporate diagram, don't worry. Everyone else was too.

OpenAI itself would later describe the structure as "bespoke." That's corporate language for: we built this thing ourselves and now we're going to have to explain it to lawyers for the next six years.

Then Microsoft arrived with billions of dollars and giant Azure supercomputers.

Now we had a nonprofit controlling a capped-profit company funded heavily by one of the largest corporations on Earth while trying to build artificial general intelligence for the benefit of humanity.

Nothing weird could possibly happen here.

Season Two: Sam Gets Fired

November 17, 2023. OpenAI's board fires Sam Altman. Not "Sam has decided to pursue other opportunities." Fired.

The board says Altman was not consistently candid in his communications and that it no longer had confidence in his leadership.

Greg Brockman gets removed as chairman of the board. Then Brockman quits.

Mira Murati becomes interim CEO. Microsoft discovers that the company it has poured billions into has just launched its CEO through the airlock.

And suddenly the entire technology industry spends a weekend watching Slack screenshots like they're election returns. Negotiations begin to bring Altman back. They fall apart. The board appoints Emmett Shear CEO. Microsoft says, essentially: fine, Sam and Greg can come work here.

Then the coup performs one of the greatest plot reversals in corporate history.

Ilya Sutskever, OpenAI's chief scientist, co-founder and one of the board members involved in removing Altman, changes sides. He publicly says he regrets participating. Mira Murati supports Altman's return. Hundreds of employees threaten to leave. The people working at OpenAI basically tell the board: you can keep the company name, we'll take the company. And that was pretty much that.

Altman comes back. Brockman comes back. Most of the board that fired Altman disappears. Total elapsed time: about five days. There are entire HBO dramas with less plot.

And Then Sam Said Something Very Interesting

When Altman returned, he wrote a note about Greg Brockman that seems considerably more important in hindsight.

He said: "Greg and I are partners in running this company." And then: "We have never quite figured out how to communicate that on the org chart."

That's an amazing sentence considering what happened next. Because over the following years, a remarkable number of the other major characters from OpenAI's early era started disappearing.

Ilya Sutskever left. Mira Murati left. Other senior researchers and executives left. Safety people left. More executives left. And now, in 2026, another substantial wave of senior people has moved on. Some departures have perfectly ordinary explanations. Health. Startups. Reorganization. Different opportunities. You shouldn't turn every resignation into evidence that somebody is fleeing the Death Star. But after enough senior departures, the departures themselves become part of the company's history.

And through all of this? Sam is still there. Greg is still there. Actually,

Greg appears to be becoming more operationally important.

Recent reporting describes Brockman reasserting himself inside the organization as OpenAI reorganizes product and infrastructure, while the company trims expensive side efforts and concentrates more heavily on revenue-producing operations.

Remember that line from 2023? Greg and I are partners in running this company. Maybe they finally figured out how to put it on the org chart.

Season Three: Everybody Leaves

This part is especially funny to me because I've written this article before. In September 2024, when Mira Murati announced she was leaving OpenAI, I wrote: "BREAKING: Sam Altman Stunned As Top Employees, Including Mira Murati, Leave OpenAI."

At the time, the question was whether we were watching normal Silicon Valley turnover or something more fundamental happening inside OpenAI.

Two years later, apparently the employee exit door still works. And again, I don't think the useful interpretation is: everyone hates Sam. That's lazy.

OpenAI itself has changed dramatically. People who joined a research organization suddenly found themselves inside one of the fastest-growing technology businesses in history. The organization they joined isn't necessarily the organization that exists today.

Which brings us to the next plot twist.

The Company That Couldn't Stop Reinventing Its Own Company

OpenAI's corporate structure has basically been a recurring subplot since the beginning.

First: nonprofit. Then: nonprofit controlling a capped-profit subsidiary.

Then: billions in outside investment. Then: board coup. Then: governance overhaul. Then: proposals to restructure again.

Eventually OpenAI concluded that the capped-profit arrangement it invented in 2019 wasn't appropriate anymore. In 2025, the company announced plans to move its commercial operations into a Public Benefit Corporation while keeping control with the nonprofit.

Later that year the recapitalization was completed, with the OpenAI Foundation retaining control and holding a massive equity stake in the commercial entity. Even Microsoft had to renegotiate the relationship. By 2026, their agreement had evolved again: Microsoft remained a major shareholder and cloud partner, but OpenAI gained considerably more flexibility to use other cloud providers and the IP relationship became less exclusive.

If you're trying to keep track at home, I'd recommend string and thumbtacks.

Why Does OpenAI Keep Doing This?

Because underneath all the drama is a very real problem. OpenAI keeps outgrowing itself.

The original nonprofit was built for research. Then the research needed billions in compute. So OpenAI created a commercial structure. Then ChatGPT happened.

Suddenly the weird research lab had hundreds of millions of users and one of the fastest-growing consumer software products in history.

Then enterprise arrived. Then agents. Then enormous infrastructure commitments. Then increasingly capable autonomous systems.

Now OpenAI isn't merely building models. It's building products. Data centers. Enterprise platforms. Developer infrastructure. Agents. Cybersecurity systems. Consumer devices and interfaces. And potentially pieces of the underlying physical infrastructure required to run the whole damn thing.

The corporate structure designed for the last version of OpenAI keeps encountering the next version of OpenAI. So they change it. Again. And again. And again. Which might actually explain much of the personnel churn too.

OpenAI isn't just scaling. It's undergoing repeated organizational phase changes.

Except Now the AI Has Joined the Cast

And here's where this season gets considerably stranger. While all of this corporate restructuring has been happening, the thing OpenAI actually builds has become dramatically more capable.

OpenAI is now publicly documenting incidents involving models behaving in unintended ways during testing. Recent disclosures have included models concealing mistakes, obtaining unauthorized credentials, uploading information publicly and finding ways to communicate across supposedly isolated environments.

Think about how strange that sentence would have sounded during the 2023 board fight.

Back then, the philosophical argument was largely about how to govern AI before increasingly capable systems arrived. Now the AI systems are becoming capable enough that they're generating their own incident reports. So OpenAI has an unusual problem. The company responsible for governing increasingly powerful artificial intelligence has spent much of its own history repeatedly redesigning how it governs itself.

That's almost too perfect.

Which Brings Us Back to Sam

So, is Sam Altman in trouble again? I went looking. At the moment, I don't see credible evidence of another November 2023 style board coup.

Actually, the current picture may point in almost the opposite direction.

Sam Altman remains at the center of OpenAI. Greg Brockman, the guy Altman explicitly called his partner in running the company, appears increasingly central to product and infrastructure.

Meanwhile, many of the other people who defined earlier eras of OpenAI are gone. That's interesting. Not necessarily sinister. Not necessarily another coup. Just interesting. Because after everything that's happened, the enduring center of OpenAI looks increasingly like the same two guys who walked back into the building together after the board tried to remove one of them three years ago.

Sam and Greg.

Maybe this isn't another rebellion. Maybe this is consolidation.

The Soap Opera Isn't a Side Story

And that's probably the larger point.

It's tempting to treat OpenAI's corporate weirdness as gossip surrounding the technology. I'm not sure we can anymore. Because how OpenAI is organized determines who controls enormously consequential AI systems.

Who raises the money. Who buys the compute. Who builds the infrastructure. Who decides what gets released. Who handles safety incidents. Who answers to regulators. And ultimately who decides what happens as these systems become dramatically more powerful.

So yes. The corporate history is entertaining. The five-day CEO coup was completely insane. The revolving door of executives is fascinating. The repeated restructuring could keep an entire generation of corporate attorneys employed. And the cast of characters occasionally behaves like someone accidentally merged Succession with Silicon Valley. But behind the entertainment is a serious reality.

One of the organizations most aggressively trying to build artificial general intelligence has spent its entire existence trying to figure out what kind of organization it needs to be in order to do it. Nonprofit. Capped-profit. Microsoft partner. Consumer technology company. Enterprise platform. Public Benefit Corporation. Infrastructure company. Whatever comes next.

Maybe that's OpenAI's real corporate history. It's not a company following a plan. It's a company repeatedly discovering that the thing it just became isn't big enough for the thing it's becoming next. And every couple seasons somebody gets fired. Or quits. Or comes back. Or restructures the entire company.

So no, I don't know whether Sam Altman is in trouble again. But I've covered OpenAI long enough to know one thing: never skip an episode.

Sources

OpenAI — “OpenAI Announces Leadership Transition”November 17, 2023 https://openai.com/index/openai-announces-leadership-transition/

OpenAI — “Sam Altman Returns as CEO, OpenAI Has a New Initial Board” November 29, 2023 https://openai.com/index/sam-altman-returns-as-ceo-openai-has-a-new-initial-board/

OpenAI — “Review Completed & Altman, Brockman to Continue to Lead OpenAI” March 8, 2024 https://openai.com/index/review-completed-altman-brockman-to-continue-to-lead-openai/

OpenAI — “Why OpenAI’s Structure Must Evolve to Advance Our Mission” December 27, 2024 https://openai.com/index/why-our-structure-must-evolve-to-advance-our-mission/

OpenAI — “Evolving OpenAI’s Structure” May 5, 2025 https://openai.com/index/evolving-our-structure/

OpenAI — “Built to Benefit Everyone”October 28, 2025 https://openai.com/index/built-to-benefit-everyone/

OpenAI — “Our Structure” https://openai.com/our-structure/

OpenAI — “The Next Chapter of the Microsoft–OpenAI Partnership” October 28, 2025 https://openai.com/index/next-chapter-of-microsoft-openai-partnership/

OpenAI — “The Next Phase of the Microsoft OpenAI Partnership” April 27, 2026 https://openai.com/index/next-phase-of-microsoft-partnership/

WIRED — “OpenAI Creates a New Framework to Disclose Bad AI Behavior” September 2026 https://www.wired.com/story/openai-releases-new-policy-for-reporting-incidents-of-model-misalignment

Previous richwashburn.com Coverage

“Altman’s Leap to Microsoft Marks a New Chapter in AI” November 2023 https://www.richwashburn.com/post/altman-s-leap-to-microsoft-marks-a-new-chapter-in-ai

“Microsoft and Apple Part Ways with OpenAI as Sam Altman Aligns with Military Industrial Complex” July 2024 https://www.richwashburn.com/post/microsoft-and-apple-part-ways-with-openai-as-sam-altman-aligns-with-military-industrial-complex

“BREAKING: Sam Altman Stunned As Top Employees, Including Mira Murati, Leave OpenAI” September 2024 https://www.richwashburn.com/post/breaking-sam-altman-stunned-as-top-employees-including-mira-murati-leave-openai

---

Rich Washburn is a technologist, strategist, and Founder & Chief AI Architect of ARIA AI Labs, working at the intersection of AI, infrastructure, communications, and capital. He also serves as Managing Partner and Chief AI Officer at Eliakim Capital.