OpenAI has unveiled ChatGPT Work, a new AI agent integrated into its flagship ChatGPT platform, designed to autonomously handle complex workplace tasks. This innovative tool utilizes the latest GPT-5.6 model to connect with various services like Gmail, Google Calendar, Slack, and GitHub. It breaks down intricate assignments into manageable steps and generates complete documents, reports, and even websites, moving beyond simple chat responses.
ChatGPT Work operates on a cloud-based virtual machine, ensuring constant availability for users without requiring their devices to be powered on, unlike some competitor solutions. This functionality allows the agent to perform tasks directly on OpenAI's servers, enhancing flexibility and accessibility.
According to product manager Tay Geri, who was involved in the development of ChatGPT Work, the aim is to democratize the capabilities of AI agents that were traditionally limited to internal corporate tools like Codex. The agent not only delivers a comprehensive outcome but also can decompose it into individual tasks to execute over extended periods.
One notable use case discussed by Geri involves managing work calendars and communications. ChatGPT Work can analyze a user's schedule, identify conflicting meetings, and determine which events require preparation. After reviewing the user's instructions, it can decide to reschedule, confirm, or cancel appointments. Additionally, users can customize writing styles and manage project workflows, creating interactive websites directly within the platform.
Geri further illustrated the tool's capabilities with an example of organizing internal product tests. Instead of manually coordinating dozens of team members, ChatGPT Work accessed information through Slack, GitHub, and existing documentation, identified participants, and organized ten separate testing sessions. Tasks that previously took months can now often be completed in just a week, including analyzing user churn and formulating product enhancement proposals.
OpenAI emphasizes its commitment to data security with ChatGPT Work, which accesses sensitive corporate information such as emails, Slack messages, calendars, and GitHub development history. Users maintain control over connections and privacy settings, and corporate accounts utilize existing security mechanisms, including the non-use of personal conversations for future model improvements.
Similar advancements are being pursued by Anthropic with its Claude Cowork and Microsoft with Copilot Cowork, as all three aim to evolve from simple chatbots to fully operational digital assistants that perform tasks through connected services. OpenAI is banking on the widespread adoption of ChatGPT, reporting 900 million weekly users and over 9 million corporate clients.
ChatGPT Work will initially roll out to Pro, Enterprise, and Edu users, followed by Plus and Business subscribers. This strategic move positions OpenAI to strengthen its foothold in the competitive AI workspace market, potentially reshaping how companies approach task automation and digital assistance.
Informational material. 18+.
