OpenAI plans to almost double its headcount this year, FT says - Fortune

OpenAI Plans to Nearly Double Its Workforce — Even as Altman Said It Would Slow Hiring

Two months ago, Sam Altman was telling developers that AI would let OpenAI "dramatically slow down" how quickly it hires. The company would do more with fewer people, he said in a January town hall. That message appears to have been superseded.

The Financial Times reported Saturday that OpenAI plans to grow its workforce from roughly 4,500 to 8,000 employees by the end of 2026 — a net addition of 3,500 people in approximately ten months. The report cited two people with knowledge of the matter. OpenAI did not respond to requests for comment from Reuters, which corroborated the story.

That is not a company hitting the brakes on hiring. That is one of the fastest headcount expansions in tech.

The Altman Contradiction

At a January 27 livestreamed town hall event, Altman laid out his philosophy clearly: "We are planning to dramatically slow down how quickly we grow because we think we'll be able to do so much more with fewer people." He framed the alternative — hiring aggressively then having to cut — as the wrong approach.

Business Insider reported on those remarks at the time. The framing made sense in context: OpenAI had just navigated internal code-red pressure from Gemini 3, was burning cash at a rate that made its $840 billion valuation feel speculative, and Altman had reason to communicate stability.

The FT's reporting, if accurate, suggests either the strategy shifted sharply in the intervening weeks, or the January comments were not a reliable signal of operational plans. It's worth noting that 8,000 employees is still a lean number for a company at this valuation — Google employs roughly 180,000 people. But adding 78 percent more headcount in a single year is not "dramatically slowing down."

What the Hiring Plan Actually Tells Us

The breakdown of where the new hires go matters more than the headline number. The FT says most will go into product development, engineering, research, and sales. That distribution is unsurprising for a company trying to turn a research lab into an enterprise software business.

What is more revealing is the new category: OpenAI is ramping recruitment of specialists focused on what it calls "technical ambassadorship" — roles designed to help corporate clients understand and deploy its tools. That is enterprise sales infrastructure. Not research. Not safety. Not compute. A corps of people whose job is to get Fortune 500 companies to use ChatGPT.

This is the company telling you something about where it thinks revenue growth comes from in 2026 and 2027. It is not from the next model release. It is from convincing businesses already interested in AI to actually build their workflows around OpenAI's products rather than Microsoft Copilot, Google Workspace AI, or a self-hosted alternative.

The Valuation Math

OpenAI closed a $110 billion funding round in February, with SoftBank, Amazon, and Nvidia among participants, valuing the company at $840 billion. To justify that number, OpenAI needs enterprise revenue at a scale it has not yet demonstrated.

Hiring 3,500 people — with a meaningful portion in sales and technical support — is a bet that the gap between enterprise interest in AI and enterprise AI deployment can be closed by human-touch onboarding. It is an expensive bet, but it is a legible strategy.

The alternative read is less flattering: that OpenAI is burning capital on headcount at the exact moment it was telling employees and the public that it had figured out how to be more efficient with fewer people. If that is the case, the January messaging was either aspirational or premature.

What We Don't Know

The FT's sources are anonymous and unverifiable from outside. OpenAI declined to comment to Reuters. There is no public confirmation of the 8,000 target from any named company official.

There is also an open question about how these numbers are defined. A 2025 public estimate placed OpenAI's headcount near 5,328 full-time roles — higher than the 4,500 baseline the FT uses. Methodology differences (FTEs versus contractors, global versus US-only) can move these numbers substantially.

The FT reporting should be treated as directionally credible — the outlet has a strong track record on OpenAI coverage — but the specific numbers deserve scrutiny until OpenAI confirms them.

The Signal That Matters

OpenAI is building an enterprise sales organization at scale. Whether the final headcount is 7,000 or 9,000 is less important than the shape of what it's building: a company that increasingly looks less like a research lab and more like a B2B software company with very expensive infrastructure.

Sam Altman saying AI would slow hiring while simultaneously planning to nearly double headcount is not necessarily a contradiction — if the new hires are doing things AI cannot do yet, namely trust-building, integration work, and enterprise hand-holding at the relationship level. But it is a tension worth watching. The narrative about AI making companies leaner does not quite fit a company planning to add 3,500 people in one year.

Sources: Financial Times (paywalled); Reuters; Business Insider report on January 2026 town hall remarks.