OpenAI opened its own enterprise records: agents now write two thirds of what companies generate
A paper posted to arXiv on 12 August by OpenAI researchers with academics from Berkeley and Wharton analyses ChatGPT Enterprise account records through March 2026 — over 1,500 organisations and more than 17 million messages at the six-month adoption mark. Output tokens across the sample rose sevenfold between June 2025 and March 2026. As of June, Codex accounted for 64% of the combined Codex and ChatGPT output tokens among enterprise customers. The widest gap is between firms: the top 10% by usage each month — the authors call them frontier firms — produce 8.3 times as many output tokens per active user as a typical firm. Adoption skews to large listed US companies with heavy R&D spending, and usage intensity is highest among early-career workers.
Related stories
Machine learning read the shape of sick brain cells and picked out nine already-approved drugs that calmed them down2026-08-21DeepSeek's cheap workhorse can now see — and on agent tasks that need eyes it says it is close to Anthropic's best
2026-08-21Given four hours and a GPU to improve the way AI is trained, the best agent scored 0.25 out of 1 — and most never tried
2026-08-21ChatGPT can now read your iMessages and send them — and the setting that lets it skip asking is the one OpenAI warns about
2026-08-21The agent invented a second person to vouch for its code. A 24-year-old in Texas refused to believe either of them.
2026-08-21