OpenAI has launched ChatGPT Work, a powerful new agent explicitly designed to automate complex, multi-step B2B workflows across enterprise applications and desktop environments. Powered by OpenAI’s newly released frontier intelligence model, GPT-5.6, and integrated directly with Codex technology, ChatGPT Work moves beyond standard conversational QA to independently execute hours-long strategic projects. It breaks overarching goals down into manageable tasks like generating dynamic spreadsheets, slide presentations, and cross-platform marketing campaigns. The ecosystem introduces “Scheduled Tasks” for automated background routines such as converting real-time communication feeds from Slack or Microsoft Teams into updated documentation and a native “Computer Use” capability that allows the agent to securely interact with local desktop tools and web browsers.
Also Read: OpenAI Retracts SWE-Bench Pro Recommendation Following Flawed Benchmark Quality Audit
Early enterprise testing across major organizations like NVIDIA, Zapier, RingCentral, and Virgin Atlantic demonstrates massive operational efficiencies, condensing intensive manual cross-app tracking and long-term data analysis from several weeks to just hours. Furthermore, OpenAI is rolling out “Sites” in public beta, letting enterprise teams quickly convert raw project data into shared, interactive web apps and live dashboards. Available immediately for Pro, Enterprise, and Edu tiers, ChatGPT Work includes robust security architecture, granular administrative controls, and customizable spend management tools to ensure safe, scalable implementation across corporate infrastructures, while the legacy Codex tool officially merges into this new unified desktop application.
