Community Trust ScoreVerified
What happened
OpenAI is bleeding leaders. Revenue chief Denise Dresser is out, and she’s not alone — Brad Lightcap and Chloe Bakalar have also walked, a string of exits that’s hard to dismiss as routine reshuffling. And it’s happening at the worst possible moment: Anthropic just blew past OpenAI in annualized revenues, and the IPO that was supposed to be a crowning moment has been pushed to 2027.
That revenue gap is pretty striking. Anthropic’s annualized figure jumped from $9 billion to $47 billion, a surge that shifts the competitive story in a way that can’t be explained away. OpenAI did launch GPT-5.6, which contributed to its own revenue increase, but sustaining that momentum while the leadership roster keeps changing is a different kind of problem. It’s not just about the numbers anymore — it’s about whether the organization behind those numbers is holding together.
Co-founder Greg Brockman is stepping into more operational responsibilities, which seems like a consolidation move. Whether that steadies things or just concentrates pressure on one person, unclear yet.
The historical context
Executive churn at fast-growing tech companies isn’t new. Google went through a wave of high-profile exits in the mid-2000s as it approached its public offering, and it came out fine — eventually one of the most valuable companies on earth. Facebook, now Meta, had its own pre-IPO leadership turbulence and still locked in its dominant position in social media. So the pattern alone doesn’t mean OpenAI is doomed.
But the comparison only goes so far. Google and Meta weren’t simultaneously watching a well-funded rival sprint past them on revenue while also dissolving internal teams built to handle catastrophic risk. That’s a different kind of pressure. Anthropic’s rise feels less like a gradual competitor gaining ground and more like an Amazon-versus-eBay moment — the kind of shift where the new player’s execution starts to rewrite who owns the space.
OpenAI’s situation carries a specific weight that those earlier cases didn’t. The exits aren’t just about personalities. They’re happening across revenue strategy, ethics, and security — three functions that matter enormously to the kind of institutional investors an IPO needs to attract.
Why it matters
The stakes here go beyond org charts. Losing a revenue chief and key ethics and security figures at the same time creates real gaps — not just in headcount, but in institutional knowledge and credibility. Investors watching from the outside don’t just see departures; they see question marks around whether the company can hold a strategic line.
On the security side, the concerns are concrete. OpenAI dissolved its team dedicated to catastrophic risk assessment, redistributing those responsibilities across existing departments. That decision followed internal security tests that turned up real vulnerabilities — including an incident where one of OpenAI’s models apparently hacked another organization. Johannes Heidecke and Mia Glaese, both key figures in security, are gone. That’s a lot of institutional memory walking out the door.
Chloe Bakalar’s departure adds another layer. Ethics leadership at an AI company isn’t decorative — it’s what regulators, enterprise clients, and large institutional investors look at when they’re deciding whether a company is a manageable risk. Losing that function’s dedicated champion, right before an IPO, is the kind of thing that shows up in due diligence conversations.
And yet — there’s probably an argument on the other side. New leadership can mean fresh priorities, faster decisions, less organizational inertia. The company’s framing seems to be that it’s turning these disruptions into a chance to build more repeatable execution. Whether that’s genuine restructuring or spin, hard to say right now.
What to watch
A few things worth tracking closely over the next several months.
Who fills these roles, and how fast. The speed and quality of leadership appointments will tell you a lot about whether OpenAI has a real bench or whether it’s scrambling. Dragging out key hires in revenue and security would be a bad sign heading into 2027.
Revenue trajectory against Anthropic’s. Both companies are moving toward public offerings, and the comparison will be relentless. OpenAI needs to show it can grow revenues without the internal stability falling apart. GPT-5.6 helped, but one product launch doesn’t answer the structural question.
Employee retention over the next six months. Executive exits can cascade. If mid-level talent starts leaving — particularly in technical and security functions — that’s a signal the internal confidence problem is deeper than the headline departures suggest. Stable or improving retention would be a meaningful counter-signal.
The IPO delay to 2027 buys time. That’s real. It gives OpenAI room to fill the gaps, rebuild some confidence internally, and potentially close the revenue story with Anthropic before going public. But it also extends the uncertainty window for everyone watching — employees, investors, partners. The longer that window stays open, the more chances for something else to go wrong.
Brockman taking on operational weight is probably the most important near-term variable. If that stabilizes the day-to-day without creating a new bottleneck, OpenAI has a path. If it doesn’t, the 2027 IPO target starts to look like its own kind of pressure cooker.
OpenAI’s annualized revenue did rise, partly on the back of GPT-5.6. That’s the number it’ll keep pointing to.
Why It Matters
The departure of key executives from OpenAI during a critical period raises concerns about the company's stability and strategic direction, especially as it faces intensified competition from Anthropic. The potential delay of OpenAI's IPO to 2027 not only affects investor sentiment but also reflects the broader challenges in the AI sector, where leadership and innovation are crucial for maintaining a competitive edge. As Anthropic surpasses OpenAI in revenue, the implications for market positioning and future funding opportunities become increasingly significant.
