OpenAI Plans to Nearly Double Workforce to 8,000 as Claude Code Gains Ground

OpenAI wants to hire 3,500 people by the end of 2026.

That would take the company from 4,500 to around 8,000 staff — one of the most aggressive hiring pushes Silicon Valley has seen in years. The Financial Times broke the story, citing people close to the company.

Why now? Because Anthropic is eating their lunch.

The coding gap

Claude Code, Anthropic's AI programming assistant, now generates more than $2.5 billion in annualised revenue. It accounts for nearly a fifth of Anthropic's business. OpenAI's equivalent product, Codex, sits at just over $1 billion.

That $1.5 billion gap matters. Coding is one of the few areas where enterprises actually pay serious money for AI tools. Software engineers — the people who spent the last decade thinking they were automation-proof — are now watching AI do their jobs. And they're paying for the privilege.

First to market is worth a lot. We had that with ChatGPT.

— Sam Altman, WIRED interview

ChatGPT launched in November 2022 and broke every growth record in tech history. But that consumer success never translated to enterprise coding. OpenAI supplied models for Microsoft's GitHub Copilot, then let the dedicated coding effort wind down. They got distracted.

Code red

In December, Altman reportedly issued a 'code red' directive internally. Google had just unveiled Gemini 3. Teams were told to pause non-core projects and redirect toward accelerating development.

But Google might not be the real threat. Anthropic — founded by former OpenAI researchers — has been focused on coding from day one. Their Claude Sonnet 3.5, released in June 2024, surprised everyone with how well it handled code. A startup called Cursor built an AI coding tool around it and saw usage explode.

OpenAI tried to buy Cursor. The founders said no. They could see where the market was going.

What Altman is betting on

The hiring will span research, engineering, and enterprise sales. That last one is new. OpenAI never prioritised enterprise sales before — they assumed the product would sell itself.

In January, Altman told employees the company would slow hiring because AI was making teams more productive. Two months later, he's planning to nearly double headcount. Something changed.

The company is also building what internal documents call a 'superapp' that merges ChatGPT and Codex into a single platform. If Anthropic's advantage is focus, OpenAI's response is integration.

It's going to be a huge business — just the economic value of it, and then also the general-purpose work that coding can unlock. I don't throw this around lightly, but I think it's one of these rare multitrillion-dollar markets.

— Sam Altman, WIRED interview

He's also said Codex is 'probably the most likely path' to AGI. If AI can write code well enough to improve itself, the whole timeline changes.

Lessons learned

OpenAI's leadership has been unusually open about falling behind. Greg Brockman, the company's president, praised Anthropic on a recent podcast:

They trained their AI models not only on difficult coding problems from academic competitions but also on real-world problems from messy code repositories. That was a lesson that we were delayed on.

— Greg Brockman, OpenAI President

While OpenAI spent 2023 and 2024 building multimodal models — text, images, video, audio — Anthropic stayed focused on making Claude better at the thing developers actually pay for.

The money

OpenAI raised $6.6 billion last year at a $157 billion valuation. Earlier this month, they announced another $40 billion. That gives them runway to hire aggressively through 2026.

There's a complication: OpenAI is still transitioning from nonprofit to for-profit. Regulators are watching. Some employees are uncomfortable with how far the company has drifted from its original mission.

Anthropic raised $8 billion in recent rounds. They can match OpenAI's hiring pace if they want to.

Why humans still matter

There's something ironic about an AI company doubling its workforce. If the technology works, shouldn't you need fewer people?

Not really. Enterprise AI requires more than a good model. You need product engineering, customer support, security expertise, and people who can actually get the thing working inside a Fortune 500 company. Anthropic understood this early. They invested in making Claude Code work in real development environments, not just on benchmarks.

OpenAI is playing catch-up. By end of 2026, they hope to have the people to compete across research, product, and sales at the same time.

What happens next

Both companies are racing toward more capable AI systems. The stakes go beyond quarterly revenue. If AI coding agents get good enough to improve their own code, development speeds up in ways nobody can predict.

For now, the fight is simpler: win enterprise customers, keep top researchers, ship products that make money. OpenAI's answer is to hire 3,500 people.

Hiring plans are expected to be finalised by Q2.