OpenAI’s Altman Concentrates Power on Path to $157 Billion Valuation

The OpenAI of today bears little resemblance to its earlier self. Since sparking an AI frenzy in late 2022 with the release of its ChatGPT chatbot, OpenAI has been rolling out and selling ever-more sophisticated AI services that can write, speak and draw to a growing number of businesses and consumers. In the process, it has also become a corporate behemoth. On Wednesday, OpenAI closed one of the largest fundraising rounds in history, valuing the company at $157 billion, up from $86 billion a little more than six months ago.

But as OpenAI races to build a large, global business, its foundation has been repeatedly shaken. The company has been roiled by an exodus of top executives and much of the brain trust that put the startup on the map, including most recently the exit of Chief Technology Officer Mira Murati. Employees at various levels of the company are now wrestling with exhaustion and concerns are simmering about OpenAI’s original commitment to responsible AI development.

Much of the shift has taken place in the 11 months since Sam Altman was briefly fired and rehired as OpenAI’s chief executive officer. Altman’s ousting followed tensions with the board over balancing the need to ensure the safety of AI models with the pressure to compete aggressively in selling software, among other issues. After Altman was reinstated, the upper ranks of the company were remade. The board he clashed with was overhauled; the C-suite that supported him was significantly expanded; and Altman is now one of the last of OpenAI’s many founders left at the startup.

OpenAI has discussed giving Altman a 7% equity stake in the company and restructuring to become a for-profit public benefit corporation, Bloomberg recently reported, though the startup’s board said it has not discussed specific numbers. With these moves, OpenAI is on the verge of becoming a more traditional tech company that builds and monetizes software – and Altman is on the cusp of becoming an even more powerful CEO.

“A lot of the opposition to him is gone, at least from what one can see from the outside,” said Alnoor Ebrahim, a professor of management at Tufts University.

In a statement, an OpenAI spokesperson said the company is “deeply committed” to its mission. “We recognize that evolving from an unknown research lab into a global company that delivers advanced AI research to hundreds of millions of people in just two years requires growth and adaptation,” the spokesperson said.

In the months after Altman’s return, four of the remaining OpenAI co-founders — originally, there were 11 — stepped back. That included Ilya Sutskever, OpenAI’s chief scientist, who played a key role in the startup’s ascent and in pushing Altman out as a member of the board that voted to fire him. Sutskever has since launched a startup focused on safe development of advanced AI systems, with no imminent plans to sell services — unlike OpenAI.

As of October, the only founders who were still actively working at the company were Altman and researcher Wojciech Zaremba, who hyperbolically likened the recent wave of departures to “the hardships parents faced in the Middle Ages when 6 out of 8 children would die prematurely.” A third founder, president Greg Brockman, is on leave until the end of 2024.

Along with the slew of organizational changes, burnout has become a key issue for some employees, according to people familiar with the matter, who asked not to be named in order to discuss private conversations. The company is looking into how to alleviate the problem — which is common at tech companies — such as by considering allowing longtime employees to take a sabbatical, one of the people said.

Brockman echoed the sentiment of feeling worn out in August when he announced his decision to take a break from work, saying in a post on X that the leave marked his “first time to relax since co-founding OpenAI 9 years ago.” Murati, meanwhile, said in a post on X in September that she was “stepping away because I want to create the time and space to do my own exploration.” A person close to Murati said she chose to do so now because she believes her team is in a strong position to keep up momentum from recent product releases.

A significant amount of turnover and change is to be expected at any fast-growing company, said Tammy Madsen, a professor of strategy and innovation at Santa Clara University. That’s especially true for a business operating in a market like AI that is rapidly evolving.

“We’d expect some of those early-stage founders to possibly move away as the company hit its hyper-growth stage because they want to create the next new thing,” Madsen said.

OpenAI has added millions of paying users and business customers, recently introduced a new model with reasoning-like capabilities and just raised $6.6 billion to help it compete against tech giants in the costly race to build more advanced AI systems.

The startup’s fast growth can be seen in OpenAI’s headcount, which now stands at roughly 1,700 workers, up from the roughly 770 it had in November when Altman was fired and reinstated. It’s also evident in the company’s leadership ranks.

Before Altman’s ousting, OpenAI had a fairly small C-Suite, including Brockman, Murati and chief operating officer Brad Lightcap, as well as Altman and Sutskever. That group has been expanded since then, with the hires of a new chief financial officer, chief product officer and chief commercial officer — roles that suggest the company is focusing more on expanding its product lineup, selling software to businesses and raising capital.

Altman is personally focusing more on the product and technical sides of the business. After Murati announced her departure, Altman said he planned to flatten OpenAI’s structure by having a number of high-ranking employees in technical roles report directly to him. At an all-hands meeting last week, Altman told employees the company is not looking to replace Murati in the CTO role for now, according to a person familiar with the matter.

The wide-ranging leadership transformation extends to the company’s board, too. All but one of the previous directors – Adam D’Angelo, the CEO of question-and-answer site Quora Inc., – have been replaced, part of OpenAI’s agreement with Altman to reinstate him.

OpenAI has enlarged and diversified its board, bringing on seven new members with backgrounds ranging from business to government, including Bret Taylor, former co-CEO of Salesforce Inc.; Larry Summers, former US Treasury secretary; and Fidji Simo, CEO of Instacart. Altman also rejoined the board in March after a probe cleared him of any wrongdoing.

The most controversial changes relate to the company’s safety teams, which have been hit by restructuring and staff departures.

In May, for instance, OpenAI decided to effectively dissolve a team called “superalignment” that focused on ensuring the safety of possible future ultra-capable AI systems. Co-founder Sutskever and another OpenAI veteran, Jan Leike, had been tasked with running the team — but Leike resigned shortly after Sutskever. Leike was publicly critical of OpenAI, saying it had put “shiny products” ahead of “safety culture and processes.” He subsequently took a similarly focused job at Anthropic, a rival company formed by early OpenAI employees that has long touted its focus on safety. A number of other team members also departed.

“Today, that team no longer exists; its leaders and many key researchers resigned after struggling to get the resources they needed to be successful,” William Saunders, a former member of the superalignment group, said during a US Senate committee hearing in September. An OpenAI spokesperson said Saunders’ comment is inaccurate, and that it is continuing to conduct safety research targeting different timelines.

The company integrated the superalignment group more deeply across its research efforts to help it achieve its safety goals, OpenAI previously told Bloomberg News, and named co-founder John Schulman as the scientific lead for OpenAI’s alignment work going forward. Months later, Schulman left, too. Like Leike, Schulman joined Anthropic, saying on social media that he would focus on safety work there.

OpenAI also unified much of its technical safety efforts — including the “preparedness” team, which was focused on “catastrophic” AI risks — under one unit called “safety systems,” as part of a broader push to ensure this work is done both across OpenAI and within a dedicated group, the spokesperson said.

OpenAI directors Zico Kolter and Paul Nakasone, members of the board’s independent safety and security oversight committee, said in a statement that their “close collaboration” with the company's safety and security teams on the release of its recent o1 AI model “demonstrated the rigorous evaluations and safety mitigations” the company puts into “every step of model development.” The board members added: “OpenAI continues to invest significantly in safety research, security measures, and third-party collaborations, and we will continue to oversee and assess their efforts.”

It’s common for fast-growing companies to rejigger their organizational footprint, Madsen said. But these particular changes raise questions about OpenAI’s commitments to developing safe AI and what had happened internally to motivate people to leave, she said.

“It is a question mark in my mind,” she said. —With Shirin Ghaffary

(Updates 25th paragraph with statement from OpenAI and C-Suite graphic to include Che Chang.)