Operator Released: OpenAI Moves from Conversational Products to Browser Task Execution
OpenAI releases Operator research preview, a model that can click, input, and complete multi-step tasks in a browser.
ChatGPT's product boundary expands from answering questions to executing tasks on behalf of users, as consumer-grade agents begin to validate real transaction and service processes.
Computer-Using Agent can now continuously perform visual recognition, reasoning, and interaction actions, but still requires frequent human confirmation.
The browser becomes the universal execution layer for agents, and e-commerce, travel, and local service platforms need to reconsider machine users.
Core metrics for To C agents should shift from usage time to number of tasks completed, time saved, and transaction conversion.
Observe payment/login scenarios, anti-automation strategies, success rates, and responsibility boundaries.