OpenAI’s London staff averaged £2.3m each in 2025, Guardian reports

OpenAI’s 110 London staff received average compensation of about £2.3m each in 2025, including salaries and shares. OpenAI UK Ltd’s accounts reportedly show £87m in salaries and £168m in shares; The Next Web said it had not independently verified the figures.

Analysis

The investable signal is labor-market pricing, not evidence of OpenAI UK Ltd’s profitability or cash burn. The reported figures are subsidiary-specific, unverified by TNW, and combine salaries with shares; without the share-award terms, accounting treatment, and scope of the accounts, they cannot be translated into cash cost, per-employee productivity, or consolidated OpenAI economics.

If the compensation is representative of scarce London AI talent, it raises the cost of building frontier-model teams and could favor firms able to recruit globally or leverage existing research infrastructure. That is a modest potential advantage for Google and Microsoft, while smaller AI developers may face higher hiring and retention costs. The second-order effect is a tighter bidding market for researchers and infrastructure talent; this does not by itself imply higher demand for compute or a change in cloud revenue.

Near term, the story is unlikely to warrant a broad AI-sector repricing. Over 1–3 months, the relevant evidence would be comparable hiring disclosures, changes in research-team growth, and whether equity compensation is recurring or tied to unusual awards. Over 6–18 months, sustained compensation inflation could pressure smaller competitors’ economics, but the article does not establish that trend. The contrarian read: a high headline figure may look like an unsustainable cost burden, yet it may chiefly reflect equity valuation and a small, highly specialized workforce. No directional trade is supported on this evidence alone.

Key Decisions for Investors

• No trade on the article alone; avoid treating the reported per-employee figure as a cash-compensation run rate or a measure of OpenAI’s consolidated cost base.
• Watch for comparable disclosures from AI labs and public technology companies, and verify award type, vesting, valuation basis, and whether figures recur before updating labor-cost assumptions.
• Use any confirmed, sustained increase in AI hiring costs as a relative watch item: it could pressure smaller model developers more than Google or Microsoft, but require evidence of hiring or margin consequences before expressing the view.
• Falsification trigger: comparable disclosures show compensation is largely one-off/noncash, or subsequent evidence indicates stable AI hiring costs and no deterioration in smaller developers’ hiring or operating economics.