Over the past year, OpenAI has endured turmoil including Musk’s lawsuit, Apple’s trade secret litigation, and executive departures. Greg Brockman, president of the $30 billion company, has quietly taken control of all core operations except research—including ChatGPT product strategy, the Codex agent platform, and enterprise sales. Although his title has remained unchanged for years, the head of API and application products has openly stated, “Everyone reports to Greg.” Brockman received approximately $10 million in equity from the Altman family office in 2017 and has never invested cash into the company. When the board abruptly ousted Altman in November 2023, Brockman stood firmly by Altman’s side and resigned together; five days later, they returned side by side. Today, nearly all co-founders have left, leaving only Altman and Brockman—now deeply aligned in interest—as the last remaining figures guiding the lab toward the eve of its IPO.

Author and source: AI New Era

New Ze Yuan report

Currently, Greg Brockman is leading OpenAI.

Altman is still CEO, but day-to-day operations are in the hands of No. 2, Brockman.

Over the past year, OpenAI has been in turmoil. Amid the chaos, one person remained steadfast, quietly accumulating power.

Ironically, Brockman’s title has remained unchanged for years: President and Co-founder.

But his influence has been growing steadily: he now controls nearly all of OpenAI’s core functions beyond research—including ChatGPT’s product strategy, the Codex unified agent platform, enterprise sales, go-to-market initiatives, and compute services.

API and product lead Thibault Sottiaux made it clear: "In the end, everyone reports to Greg."

OpenAI is in turmoil, with Brockman holding all the power.

This year, OpenAI had a disastrous start.

First, it was embroiled in lawsuits.

Engaged in a high-profile courtroom battle with co-founder Musk, lasting several months.

又被 Apple filed a high-profile trade secret lawsuit.

By July, an unreleased model had escaped containment during security testing, infiltrated Hugging Face, and OpenAI was forced to publicly acknowledge its mistake, sparking widespread public outcry.

As the IPO threshold approaches, executives are successively stepping down.

Throughout the year, key figures have departed: Sora head Bill Peebles, VP of Science Kevin Weil, and B2B Applications CTO Srinivas Narayanan—all left within a single month. Two others stepped down for health reasons: CMO Kate Rouch and AGI Deployment CEO Fidji Simo. Simo has been on medical leave since April.

The storm hasn't subsided, but April is the true turning point.

This month, CRO Denise Dresser, who had been under heavy pressure since the April turmoil, suddenly stepped down after only eight months in the role; days later, Brad Lightcap, who had served as COO for many years and recently overseen “special projects,” also announced his departure, saying he wanted to “pursue new horizons.”

Ross Carmel, a partner at the law firm SRFC, told The Verge that it's not uncommon for executives to leave before an IPO, but "it's unusual for everyone to leave within such a short time window."

On August 17, Brockman personally appeared on CNBC’s Squawk Box to address the situation. He said the leadership turnover was not unusual: “At different times, we have different leadership teams... I am a constant, and Altman is a constant. It is precisely because of this resilience and diversity that we are stronger.”

Chris Malone leaves in 9 days.

This person oversees data center expansion, and computing power is precisely OpenAI’s most expensive and most important moat to clarify compared to Anthropic.

According to reports, after the infrastructure reorganization, he no longer reports directly to Brockman but instead reports to Vice President Sachin Katti. This appears more like a centralization of power followed by a downward shift in hierarchy, rather than a natural turnover of personnel in each era. Brockman’s statement on August 17 that “nothing unusual” was happening required further clarification by August 26.

Another layer of practical benefit: cutting a few of the highest executive salaries to make the financial statements look better before the IPO. “For OpenAI, which lags behind Anthropic in revenue, you either need to be more profitable or closer to profitability.”

Analysts have even provided a timeline: Before the end of September, OpenAI will launch a consumer-grade product under Brockman’s leadership, allowing him to “prove he deserves the position.”

Variables came and went in waves. At the table, only two constants remained.

The two people at the table

Brockman of OpenAI is now second only to Altman.

Before OpenAI, Brockman was best known for expanding Stripe’s business and was renowned internally for championing the company’s marketing efforts.

In December 2015, OpenAI debuted as a non-profit laboratory.

At the time, Musk was the largest early investor and public face; Altman was still president of Y Combinator, responsible for recruitment and external messaging; Brockman left his role as CTO of Stripe, and the early office was located in his living room; Ilya came from Google Brain, lending the lab academic credibility.

In 2017, OpenAI was still merely a research organization promoting itself as "nonprofit" and "saving humanity."

But in that year’s personal journal, the next intriguing remark was: “Financially, what could make me $1 billion?”

During the trial of Musk v. OpenAI on May 4, 2026, a stunning court document was made public. Brockman confirmed during the trial that the value of his current OpenAI shares is nearing $30 billion.

More strikingly, he admitted that he had never invested a single dollar of cash into the company.

During the trial, a deeply hidden financial link was revealed: as early as 2017, Altman transferred approximately $10 million in equity from his family office directly to Brockman as compensation for his "sweat and tears" work at OpenAI.

In addition, Brockman also holds significant stakes in other related startups of Altman.

This bombshell sent Musk into a rage. His fans and supporters launched a barrage of criticism, labeling Brockman a "hypocrite." Musk himself sarcastically referred to him as "one of the most shameful people in America."

Setting aside the courtroom debates, this lawsuit sends a very clear message to the outside world: Altman and Brockman have long been deeply aligned in their interests.

He also sways back and forth.

Interestingly, the two were not always united.

During OpenAI’s early, challenging years, Brockman did not always defer to Altman on major decisions regarding power dynamics and technological direction.

Sometimes, he sided with Chief Scientist Ilya Sutskever, expressing concern over Altman’s overly aggressive commercial ambitions; yet he also feared that the powerful Musk would completely overpower Altman.

By 2017 to 2018, intense power struggles erupted among a few individuals over whether OpenAI should transition to a for-profit model and how to finance computing power.

Musk demands absolute control, otherwise he will merge OpenAI into Tesla.

Brockman strongly opposed this: "No one should have complete control over OpenAI."

Ultimately, the negotiations broke down, and Musk exited in 2018.

Subsequently, in 2019, OpenAI officially established a "capped-profit" structure. Long before that, the $10 million family office equity investment had already laid the foundation for a dual contract of money and power between Altman and Brockman.

After OpenAI established its for-profit structure in 2019 and Altman officially became CEO, Brockman temporarily relinquished much of the day-to-day management authority.

But Brockman frequently appears abruptly in various projects as the founder.

He would abruptly overturn well-established plans at the last minute, inserting his own code or features. This management style—of storming into others’ projects to create chaos and leaving other executives (such as CTO Mira Murati) to clean up the mess—fueled widespread internal resentment and ignited the spark of the “OpenAI palace coup.”

But it cannot be denied that his contributions are equally significant.

Brockman’s consistent role has been to turn OpenAI’s research into something that can be monetized.

In 2020, he praised GPT-3 and the developer API: “This is the first time OpenAI has had a general AI system with immediate commercial value—it’s actually useful.”

It was Brockman who accomplished OpenAI’s dramatic transformation from a “research lab” to a “profitable business.”

In 2022, he and Altman demonstrated GPT-4 on a biology question at Gates' home; in March 2023, he hosted the live launch of GPT-4.

The real turning point

The real dividing line was on November 17, 2023.

On that day, Ilya, in conjunction with the board, abruptly removed Altman from his position as CEO, citing "lack of candor with the board." Meanwhile, Brockman, who was then chairman of the board, was completely blindsided and immediately removed from the board.

At that moment, facing the greatest crisis of his political career, Brockman did not hesitate. He posted on X: "Given today's news, I am resigning."

He immediately voiced support for Altman and joined him at the hotel to plan a new AI company. Five days later, under intense pressure from Microsoft and the threat of a near-universal employee strike, the original board collapsed, and Altman returned as king, with Brockman by his side.

On August 5, 2024, he publicly stated: "Taking a break until the end of the year—my first real rest in nine years."

Originally planned for about four months, returning early in mid-November; internally, it was said that we were discussing a new role with more technical challenges with Ultraman.

Today, very few of the original co-founders remain—

Among the co-founders, Musk has left, Ilya has left, Karpathy and Schulman have left, Murati departed long ago, and Wojciech Zaremba is no longer in a core position at OpenAI.

Only Altman and Brockman have stayed at the table from the lab all the way to preparing for an IPO.