OpenAI is paying employees $1.5 million each

OMG, how am I going to start an AI department at the Report?

OpenAI is paying employees $1.5 million each

A Wall Street Journal scoop reveals average stock-based comp of $1.5 million per employee, a staggering figure even in world-historical terms

By Catherine Baab, Quartz Media

Published Yesterday

OpenAI is paying its employees at levels that have little precedent in Silicon Valley history — $1.5 million in stock-based compensation per employee in 2025. That’s according to financial projections shared with investors and first reported by the Wall Street Journal.

Such compensation dwarfs what earlier generations of tech giants paid their workforces before going public. Adjusted for inflation, OpenAI's equity payouts are multiples above the norm for large tech firms over the past two decades, and even several times higher than what Alphabet or Meta disclosed ahead of their IPOs.

Ferocious competition for AI talent

The $1.5 million figure makes clear just how fierce the competition is for top AI talent, with compensation for individual employees beginning to resemble packages that were once reserved for founders.

As generative AI models become central to corporate strategy, researchers and engineers with the most relevant expertise are becoming some of the most expensive employees anywhere in the world. To retain current employees and win sought-after job candidates, rival research labs and tech giants must make staggering offers.

A classic example of "prisoner's dilemma"

In game theory, such a dynamic is known as “prisoner's dilemma,” which describes situations in which competition effectively forces participants to escalate costs beyond what any individual would prefer.

In this case, if leading AI firms collectively capped compensation, they'd all benefit from lower costs and less-diluted equity. But no single company can afford restraint when competitors are bidding aggressively — not when falling behind in talent means falling behind in the technology itself. The result is a bidding war none of the participants can escape, even as it pushes the economics of AI development into increasingly unprecedented territory.

Accordingly, OpenAI has had to loosen internal equity rules to stay competitive. One recent change removed a waiting period that required employees to stay six months before stock awards began vesting — a nod to a job market where researchers frequently move between companies and can command immediate payouts.

Compensation now consumes 50% of OpenAI's revenue

For OpenAI, this compensation strategy carries steep costs — in several senses. The company has about 4,000 employees at present, and stock compensation now consumes about half of the company’s annual revenue. That’s an unusually high share that likely contributes to operating losses and, at the same time, functions to dilute existing shareholders.

In historical terms, the numbers are even more jarring. Among major tech companies that eventually went public, stock-based compensation typically accounted for a modest, single-digit percentage of revenue in pre-IPO years. But at OpenAI, it now approaches 50%.

What’s more, per Wall Street Journal reporting, OpenAI’s investor materials suggest those costs will keep rising for years, even as the company continues to invest heavily in infrastructure and development.

So, can it last?

Whether the strategy proves sustainable or not, it captures something essential about the contemporary tech economy. The overall pattern is familiar — a handful of firms create world-historical value while a tiny fraction of workers capture most of the gains — but the scale makes earlier booms look quaint. OpenAI is just making the math more explicit and more extreme.

📬 Sign up for the Daily Brief