OpenAI lets an AI supervise your AI for free
OpenAI just rolled out its Auto-review feature for free to all signed-in ChatGPT users without touching plan quotas. The mode deploys a secondary agent to watch over the primary model's moves, blocking high-risk or off-track actions automatically. It basically solves prompt approval fatigue on long agent workflows.
Running autonomous AI agents usually comes with an annoying dilemma: either babysit every single step or risk a mess in the background.
OpenAI has rolled out its Auto-review feature for free to all signed-in ChatGPT users. You can enable it directly under settings > permissions.
A second agent on guard duty
Instead of making you approve every single action—which quickly causes decision fatigue—Auto-review deploys a secondary agent to watch over the main one.
Its sole job is keeping long-running tasks on track, blocking high-risk actions and filtering out anything that drifts from your original intent. You no longer have to spend time configuring tedious rule sets, and the feature does not draw any usage from your existing plan.
Why it matters
Letting AI agents work autonomously is great until you find yourself turned into a full-time human approval button.
We think making dual-agent supervision free is a smart move by OpenAI. It removes the friction of letting agents run unattended tasks without forcing users to burn through their quotas or micromanage every permission.
The real test will be reliability. If the reviewer agent shares the same blind spots as the primary model, or gets overly paranoid and blocks legitimate actions, users might still find themselves stepping in.
Sources (1)
- x.com(opens in a new tab)Originalx.com