OpenAI Launches ChatGPT Work for Enterprise Teams With Agentic Controls

Short summary

OpenAI launched ChatGPT Work, an agentic capability within ChatGPT Enterprise that executes multi-step tasks across apps and files under organizational controls. Powered by GPT-5.6, it can carry projects for hours, produce finished outputs like dashboards and briefs, and run scheduled monitoring tasks. Governance controls cover accessible context, connected tools, and actions requiring human approval, making it an extension of—not a replacement for—ChatGPT Enterprise.

• •ChatGPT Work adds agentic multi-step task execution within ChatGPT Enterprise, powered by GPT-5.6
• •Governance controls cover context access, connected tools, and human-approval gates for actions
• •Rolling out incrementally to eligible Enterprise and Edu workspaces; includes scheduled tasks and cross-app coordination

Generated with AI, which can make mistakes.