With our partner Salesforce, unify sales, marketing and customer service. Accele your growth!
Openai has released today Operatoran autonomous agent capable of navigating the web, interacting with graphic interfaces and performing tasks for its users. Based on COMPUTER-USING Agent (CUA)Operator combines GPT-4O’s vision capacities and reinforced learning to interact with websites without requiring personalized APIs.
Accessible in preview for professional users in the United States, Operator can perform repetitive tasks such as filling forms, ordering races or personalizing workflows. It works in collaboration with the user, requesting confirmations for sensitive actions and assigning control if necessary, in particular for the entry of personal or financial data.
We could not confirm your registration.
Your registration is confirmed.
Operator is based on strategic partnerships with companies such as Instacart, Doordash and Opentable, and also explores applications in the public sector, in particular with the City of Stockton to simplify access to municipal services. Despite its advanced capacities, this agent remains in the beta phase and still has limits on complex interfaces.
In terms of security, Openai has implemented control mechanisms, including supervision mode, protections against malware, and personal data management options.
Ultimately, Openai plans to integrate Operator in Chatgpt and offer the CUA model via an API to allow developers to design their own autonomous agents.
