ChatGPT rolls out refreshed UI and advanced permissions

OpenAI is addressing usability and integration friction by introducing automatic approvals, signaling a push towards more agentic workflows where constant confirmation is bypassed for trusted actions.

* Auto-approval for low-risk actions reduces notification fatigue, moving ChatGPT closer to a seamless background assistant.

* Refined permission settings and intelligence modes allow power users and enterprises to customize safety and privacy boundaries.

* The focus on connector search indicates a preparation for a much larger ecosystem of integrations and custom tools.