OpenAI Loses Two Senior Executives and Delays IPO to 2027

OpenAI entered July 2026 with its second-in-command gone and its IPO pushed back a full year. Fidji Simo, who joined the company last August to handle day-to-day operations and free up CEO Sam Altman, announced she is leaving after her chronic neuroimmune condition worsened beyond what she had originally expected. Her departure comes just months after three other top executives left on the same day in April, compounding concerns among investors and staff about the company's leadership stability heading into a planned public offering.

What happened

Data point	Detail
Fidji Simo departure	Stepping down as OpenAI’s number-two; will stay on as part-time adviser
April exits	Kevin Weil, Bill Peebles, and Srinivas Narayanan all left on the same day
IPO timeline	Delayed from 2026 to 2027
Bank of America credit line	$520 million extended to OpenAI ahead of IPO
OpenAI target valuation	More than $1 trillion
OpenAI 2025 net revenue	$13.1 billion, with a net loss of $38.5 billion
Anthropic valuation	$965 billion after a $65 billion funding round
OpenAI early-2026 funding	$110 billion from SoftBank, Amazon, and Nvidia at an $840 billion valuation

Simo joined OpenAI in August 2025 with a broad mandate: oversee the CFO and chief revenue officer, lead product and business divisions, and generally give Sam Altman room to focus on strategy rather than operations. She announced a medical leave in April after her condition flared, and this week she told staff in an internal note viewed by The Wall Street Journal that her road to recovery will be far longer than she had hoped.

The condition she has been managing since 2019 is Postural Orthostatic Tachycardia Syndrome (POTS), a chronic neuroimmune disorder that affects the autonomic nervous system. “This has been one of the hardest decisions of my career, but my body left me no choice,” she wrote to colleagues.

The three executives who departed on the same day in April were Kevin Weil (head of scientific research, previously chief product officer), Bill Peebles (who ran the now-defunct Sora video app), and Srinivas Narayanan (CTO for business-to-business applications). According to the Journal, Simo’s earlier absence had already created a leadership vacuum that unsettled investors and staff.

Why it matters

The executive churn is bad optics for a company preparing to go public, but the competitive picture is what makes it genuinely concerning. Anthropic, OpenAI’s closest rival, has now surpassed OpenAI’s valuation for the first time. Enterprise customers appear to be favoring Anthropic’s Claude assistant at this stage, and Anthropic confidentially filed for a US IPO last month, just weeks after closing a $65 billion funding round at a $965 billion valuation.

OpenAI’s own financials are complicated. According to a report cited by Moby, the company generated $13.1 billion in net revenue in 2025 but posted a net loss of $38.5 billion. Most of that loss stems from accounting changes tied to its conversion from nonprofit to for-profit status, which pushed early investor obligations onto the books as debt. Strip that out and the operating picture looks closer to the $5 billion loss reported for 2024, though neither figure is reassuring for a company targeting a $1 trillion valuation.

Bank of America’s $520 million credit line, reported by Reuters this week, is the bank’s first direct loan to OpenAI. Reuters noted that BofA is positioning itself for advisory roles on both the OpenAI and Anthropic IPOs, which says something about how seriously the financial world is still taking both companies despite the turbulence.

Our earlier coverage of OpenAI’s head of safety leaving noted a similar pattern: key people exit, teams get restructured, and the public justification rarely tells the whole story.

Our take

Simo’s departure is genuinely sympathetic on a human level. POTS is a serious, often debilitating condition and her statement makes clear this was not a strategic move. But the timing matters regardless of the reason. OpenAI now heads into its IPO process without the executive who was specifically hired to manage the business so Altman could focus elsewhere. That gap is real and it is not yet clear who fills it.

The Anthropic valuation flip is the bigger story hiding inside this one. A year ago, “OpenAI vs. everyone else” was not much of a competition in enterprise software. Now Anthropic is valued higher, filing for IPO, and winning enterprise deals. If you are a business currently evaluating AI vendors, that race is worth watching before signing a long-term contract with either provider.

For businesses already building on OpenAI’s APIs, none of this week’s news changes anything immediately. But if you are planning an AI integration project that will depend heavily on a single provider, now is a reasonable time to pressure-test that assumption and at least prototype against a second model.

What to do about it

1. Audit any OpenAI API dependencies in your current stack and note which ones are provider-specific versus easily portable.
2. Run a small test workload through Anthropic’s Claude API to establish a baseline comparison on cost and output quality for your use case.
3. Watch the IPO filing disclosures when they become public. Prospectus documents will contain financials and risk factors that are far more detailed than anything currently reported.
4. If you are evaluating enterprise AI contracts, push vendors for pricing that does not lock you in for more than 12 months while the competitive landscape settles.

The IPO race between OpenAI and Anthropic will produce more disclosure than the AI industry has ever had to make. That transparency, whenever it arrives, will be more useful than any press release either company has issued.

Source: Bing News · Sora (AI video)