0
Applied AI·November 21, 2026·1 min read

OpenAI says its agents now finish nearly half of 2 to 4-hour research tasks

Share

If agents can autonomously complete 46% of 2–4 hour research tasks — up from 23% in eight months per OpenAI’s own report — the bottleneck moves from execution to scoping and verification. Teams that standardize task templates, acceptance criteria, and review workflows will capture this compounding curve; everyone else will just generate more untrusted drafts.