Dots Post-Launch: Permission Fatigue and User Complaints

## Dots Post-Launch User Complaints (Oct 26-27, 2026) **Source**: https://www.techcityauthority.com/2026/10/chatgpt-dots-users-permission-prompts-lost-approvals-codex-usage.html ### Reported Issues - Paying users say Dots keeps asking for permissions repeatedly - Users report the agent forgets approvals already given - Agent stalls on tasks it should be able to complete - Some users describe getting stuck in permission loops ### Verification Status - **Anecdotal only**: TechCityAuthority found only a small cluster of posts - **Not independently verified**: The outlet could not verify the complaints independently - **No official response**: No OpenAI response to these complaints was found - **Low confidence** on specific claims; medium confidence that some friction exists given the nature of approval-based agent design ### Context - Dots launched as OpenAI's "always-on" agent with its own cloud computer - Requires user approvals for actions - Costs $200/month for Pro tier access - Contrast with Anthropic's Claude Duet (Oct 21 launch), which led with upfront limitations and waitlist gating ### Implications for Series - Illustrates the consumer pitfall: agents that promise autonomy but create new friction - Complements the Eon study (Nov 4) on agent reliability and honesty - Provides concrete user experience angle on "overpromise" theme planned for Nov 5 - Contrast with Anthropic's transparency approach strengthens the hype-vs-receipts frame