# OpenAI Is Testing a Codex Mode That Keeps Working Until a User Stops It

Code reviewed by WIRED describes an agent that generates its own follow-up tasks across sessions and can message a user unprompted.

- Published: 2026-08-29T06:21:15.195Z
- Canonical: https://polylog.news/ai/2026-08-29/openai-is-testing-a-codex-mode-that-keeps-working-until-a-us
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [Gizmodo](https://gizmodo.com/nevertheless-openai-persists-with-new-always-on-agent-2000804088), [Slashdot](https://features.slashdot.org/story/26/08/27/224230/openai-is-developing-a-persistent-ai-agent)

OpenAI is building a persistent mode for Codex, its coding agent, according to code that WIRED reviewed and that several outlets have since described. The setting appears in the reasoning-effort menu and lets the agent spend compute, tokens…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-08-29/openai-is-testing-a-codex-mode-that-keeps-working-until-a-us (subscription information: https://polylog.news/pricing).