ChatGPT Work: Agent for Ambitious Tasks with Codex and GPT-5.6
OpenAI identified a clear gap: ChatGPT was great at answering questions but not at doing real, multi-step work across apps and files. The tension is that users want an AI that can take on ambitious tasks—like creating a slide deck from scattered emails, reconciling a month-end budget, or building a campaign brief from raw research—without needing constant handholding. Existing chat interfaces break down when work requires hours of context, tool switching, and autonomous execution.
ChatGPT Work is the concrete answer. It is powered by Codex—already used by over 5 million people weekly, with more than 1 million using it for non-coding tasks—and the new GPT-5.6 model, which is state of the art at multi-step reasoning. The agent can connect to plugins for Slack, Microsoft Teams, Google Drive, and SharePoint, pull context via @-mentioning apps, and create finished materials like sheets, slides, docs, and web apps. It supports Scheduled Tasks for recurring work (e.g., monitoring Slack updates and refreshing an agenda each week). On desktop, it adds a built-in browser and Computer Use to interact with local files and apps. Enterprise controls include auto-review, which blocked 100% of attempted data exfiltration in red-teaming. Usage is metered similarly to Codex, with admin spend controls.
The real-world takeaway for builders is that agentic work is no longer theoretical. Inside OpenAI, nearly 100% of teams now use ChatGPT Work. Sales turned a discovery conversation into a tailored proof of concept within 24 hours (normally weeks). Finance reduced month-end close from days to hours. The article advises giving it a task you already know well—like analyzing a budget variance—to learn its strengths. For serious technical readers, the key insight is the balance between autonomy and control: the product deliberately checks in for approval on important actions and lets admins govern access to tools, apps, and local files. If you are building or buying an AI agent platform, this pattern—frontier model + tool orchestration + scheduled execution + granular security—is the blueprint for production-grade work automation.
