OpenAI’s leadership churn reached one of its most commercially important veterans on Tuesday, August 11, when Brad Lightcap, the company’s former chief operating officer and most recently its special-projects lead, said he would leave to start a new venture. Lightcap announced the move himself after roughly eight years at OpenAI.

His departure landed hours after The Financial Times first reported that Chloé Bakalar, OpenAI’s AI ethics lead, had quietly left in July, less than a year after joining. The timing matters because OpenAI was valued at $852 billion in March and said on June 8 that it had confidentially filed for an initial public offering, giving future public-market investors a more immediate reason to watch leadership continuity.

“I’ll be moving on from OpenAI to start something new,” Lightcap wrote. The old picture was of a management reshuffle in which Lightcap had stepped away from the COO title but remained close to CEO Sam Altman on special projects. The new fact is that he is leaving the company entirely, removing an executive who helped turn OpenAI from a research lab into a global commercial business.

Two exits carry different signals

Lightcap’s exit and Bakalar’s departure carry different signals. Lightcap moved out of the COO job in April, with chief revenue officer Denise Dresser absorbing much of his former remit, so OpenAI had already redistributed many of his day-to-day responsibilities. Bakalar, by contrast, was the company’s only dedicated ethicist, after previously serving as chief ethicist at Meta.

There is also an important timing caveat. The Financial Times said Bakalar left “last month” but did not disclose the date. That means her departure was revealed inside the July 28 to August 11 window, but her actual last day cannot be definitively placed inside it.

The broader pattern began earlier in July. WIRED first reported on July 7 that Joshua Achiam, OpenAI’s chief futurist and former head of mission alignment, had told colleagues he would leave after nearly nine years. Three days later, WIRED first reported that Johannes Heidecke, head of safety systems, was also leaving as OpenAI reorganized its safety and research teams.

Fidji Simo’s move was different again. The Wall Street Journal first reported on July 9 that OpenAI’s No. 2 executive would step down from her full-time role after an extended medical leave. Simo moved to a part-time advisory position, so she should not be counted as a clean break from the company.

A $852 billion company approaches public markets

For investors, the departures arrive at a sensitive point in OpenAI’s transition from private company to potential public stock. OpenAI’s June filing set no IPO date, but Reuters reported a target of up to $1 trillion. CFO Sarah Friar told CNBC in April that OpenAI plans to reserve a portion of its IPO shares for individual investors. Because OpenAI remains private, there is no public share-price reaction to use as a clean signal.

The competitive position is strong but no longer uncontested. Similarweb ranks ChatGPT No. 1 among AI chatbot and tools websites for July 2026, ahead of Gemini and Claude. Artificial Analysis currently scores Anthropic’s Claude Fable 5 at 62 versus 61 for GPT-5.6 Sol in its composite Intelligence Index, which combines tests of reasoning, knowledge, mathematics and coding.

Safety turnover raises a governance question

That makes management turnover more than an internal personnel story. OpenAI is simultaneously preparing for public-market scrutiny, defending a consumer lead, competing with Anthropic and Google on frontier models, and dealing with more demanding safety questions as AI systems become capable of acting with less human supervision. Recent OpenAI security incidents have made that final issue especially visible.

Nothing public establishes that Bakalar, Achiam or Heidecke left because of those incidents, and Lightcap said he is leaving to build something new. The departures therefore should not be read as evidence of a single internal crisis. They do, however, concentrate succession risk at a company whose valuation assumes it can scale products, safety and governance at the same time.

For prospective investors, the question is no longer whether OpenAI can attract exceptional people. It is whether the institution can keep compounding after some of the people who built its first era decide their next chapter belongs elsewhere.