OpenAI Codex agents now primary tool across all departments; 80% of users complete 30+ minute tasks
<cite index="51-3">OpenAI research released in June 2026 shows that from December 2025 to May 2026, the share of individual users making requests estimated to correspond to work taking more than 30 minutes rose to 80.6%, work over one hour to 70.2%, and work over eight hours grew fastest from a low base.</cite> <cite index="51-1">Among daily active users at OpenAI, the heaviest users ask Codex to run many hours of agent work in a single day, with users at the 99th percentile regularly generating more than 60 hours of Codex agent turns per day, distributed across multiple parallel agents.</cite>
<cite index="56-1">At OpenAI itself, Codex is now the primary AI tool across every team, accounting for more than 85% of output tokens for the average employee.</cite> <cite index="51-4">By early June 2026, non-developer individual users multiplied 137 times since August 2025, non-developer organizational users increased 189-fold, and non-developer OpenAI users increased 12-fold.</cite> <cite index="51-4">The shift does not mean every non-developer is using Codex the same way as an engineer; rather, more non-developers are using Codex for some kind of agentic work.</cite>
<cite index="57-1">OpenAI's enterprise business now makes up more than 40% of revenue and is on track to reach parity with consumer by the end of 2026.</cite> The research documents four clear trends: longer-horizon work, multi-step task chaining, cross-functional collaboration, and adoption velocity beyond engineering roles into finance, legal, recruiting, and other knowledge-work departments.
<cite index="51-3,56-1">For architects and operational leaders: this signals agentic work is no longer experimental—it's infrastructure.</cite> The speed asymmetry is stark: teams with 24/7 running agents on multi-step workflows gain compounding advantage. <cite index="53-1,53-4">Gartner projects 40% of enterprise applications will include AI agents by end-2026 (vs. under 5% in 2025), yet survey data shows daily AI users report higher engagement but lower confidence in productivity, flagging an adjustment period where workflows must be re-architected, not just augmented.</cite>
Sources
- Primary source
- Digit
“By May 2026, 81% of sampled users had asked Codex to complete at least one task over 30 minutes, 70% for over one hour, 26% for over eight hours”
- Asanify
“Gartner projects 40% of enterprise apps will include AI agents by end-2026, up from under 5% in 2025”
- OpenAI
“Enterprise now makes up more than 40% of OpenAI revenue, on track to reach parity with consumer by end-2026”