What is happening at OpenAI? Twelve executives left in a year, and the org chart may be the busiest product
Business Insider listed 12 prominent OpenAI executives and leaders who left in 2026 for reasons including startups, health, role changes, and governance disagreements. It is not proof of one feud, but it shows the organization is repeatedly redrawing its boundaries.
The bottom line
While OpenAI is preparing for the next stage of business expansion, it is like installing a revolving door: Business Insider counts that 12 important executives or core leaders will have left in 2026.
The list includes former COO Brad Lightcap, application business CEO Fidji Simo, chief revenue officer Denise Dresser, former chief product officer Kevin Weil, Sora head Bill Peebles, as well as the heads of ethics, security, robotics and enterprise business.
This could easily be written off as “an explosion of internal strife at OpenAI.” But the reasons given by public reports are mixed: some are starting businesses, some are adjusting for health reasons, some are turning to consultants, some are expressing dissatisfaction with the governance and defense agreement, and some just feel it is time to move to the next stage.
So the real story isn't "12 people got kicked out," but this: Can the organizational structure remain stable when a company is simultaneously sprinting for models, products, advertising, enterprise, and a potential IPO?
This is not one resignation; an entire organizational map is changing hands
It’s not the same kind of positions that are leaving in 2026.
- Business layer: Brad Lightcap, Fidji Simo, Denise Dresser.
- Product and R&D layer: Kevin Weil, Bill Peebles, head of Sora.
- Enterprise & Technology: Srinivas Narayanan, Barret Zoph.
- Marketing and Robotics Layer: Kate Rouch, Caitlin Kalinowski.
- Safety and ethics layer: Johannes Heidecke, Chloé Bakalar, Joshua Achiam.
What’s most noteworthy about this list is the span: from revenue and applications, to Sora, robotics, to security and ethics. It's not like a wave of departures from a single team, but more like the company is constantly redrawing "who is responsible for what."
The most easily overlooked reversal: many people leave for different reasons.
Fidji Simo publicly cited long-term health issues and moved into an advisory role; Kate Rouch left her marketing job to focus on recovery; Brad Lightcap said he was starting something new while emphasizing that he still believed in OpenAI; and Caitlin Kalinowski publicly expressed concerns about the way defense agreements are governed.
Putting all these people into the "infighting" narrative is very convenient for communication, but in fact it is not accurate at all.
But the reasons are different, which does not mean that the phenomenon is not worthy of attention. Frequent changes at the top can impact product roadmaps, customer relationships, risk judgments, and internal memory. Especially when ethics and safety positions are also being adjusted, the outside world will naturally ask: Is the company becoming more mature, or is it constantly reorganizing amid rapid expansion?
The busiest thing about OpenAI may not be the model, but the organizational structure
OpenAI has gone through several rounds of executive and researcher changes over the past year. At the same time, the company is also advancing GPT‑5.6, Codex, ChatGPT advertising, enterprise products, and new security controls.
Each thing can be explained individually, but when put together, it creates a management problem: the research team wants to chase capabilities, the product team wants to chase releases, the business team wants to chase revenue, the security team wants to chase brakes, and the board of directors also has to face scale and governance.
That's why "12 Executive Departures" is more informative than regular turnover. It does not directly prove that OpenAI is out of control, but it shows that the company has entered a stage where it relies on continuous shifting of gears to move forward.
Our verdict: Do not rush to declare OpenAI finished
Resigning does not mean failure, and staying does not mean stability. It's not uncommon for startups to replace senior executives during periods of rapid expansion; what really needs to be seen is whether positions are filled in a timely manner, whether responsibilities are clearer, and whether product and security decisions become slower or more consistent as a result.
The most valuable part of this news is not to label OpenAI as a "controversy", but to let us see the true cost of a company in the AI competition: model launches can be updated on a weekly basis, but organizational trust and governance experience cannot be upgraded with one click.
What to watch
- Whether OpenAI continues to see core security, ethics, and application job changes.
- Will the new person in charge centralize power or continue to split business lines?
- Whether executive changes affect ChatGPT, Codex, advertising and enterprise product routes.
- How companies balance revenue velocity and security governance in a potential IPO narrative.