Issue 32Published September 7, 2026
Editor noteThis archived Ash AI Daily issue retains the delivered editorial briefing and final cards.
Story 1AI Daily
OpenAI says coding agents now exceed human research labor
OpenAI reports that by mid-August its research organization was using 3.1 agent-workdays for every human workday, measured as eight-hour-equivalent runtime. It says the median researcher was using more than $600 per day of inference at API prices, and that experiments per active experimenter hit a tracking high in August. The company also says its measurements are preliminary: agents still need meaningful human steering, particularly on longer tasks.
- OpenAI reports that by mid-August its research organization was using 3.1 agent-workdays for every human workday, measured as eight-hour-equivalent runtime. It says the median researcher was using more than $600 per day of inference at API prices, and that experiments per active experimenter hit a tracking high in August. The company also says its measurements are preliminary: agents still need meaningful human steering, particularly on longer tasks.
Why it matters
This is a useful operating signal, not an independent productivity study. If the trend generalizes, research teams will need to manage agents as throughput infrastructure—setting priorities, reviewing outputs and allocating compute—rather than treating them only as individual coding tools.
Sources
OpenAI says coding agents now exceed human research laborOpenAI — Research acceleration: The view inside OpenAI · Sep 7, 2026
