OpenAI Launches ChatGPT Agent for Automatic Task Completion

OpenAI has launched a new general-purpose AI agent in ChatGPT that can complete a wide range of tasks, including navigating calendars, generating presentations, and running code. The agent combines capabilities from previous OpenAI tools, including Operator and Deep Research, and can be activated by selecting "agent mode" in ChatGPT's dropdown menu. The model underlying ChatGPT agent offers state-of-the-art performance on several benchmarks.

OpenAI has introduced a new AI agent within ChatGPT, capable of performing a wide array of tasks, from navigating calendars to generating presentations and running code. This general-purpose agent, accessible through a "agent mode" option in ChatGPT's dropdown menu, combines the strengths of OpenAI's previous tools, Operator and Deep Research. The new agent can handle complex workflows, including tasks such as analyzing competitors and creating slide decks, all within a unified interface.

The ChatGPT agent is designed to offer state-of-the-art performance on various benchmarks, leveraging the intelligence and conversational fluency of ChatGPT. It can navigate websites, filter results, and even prompt users to log in securely when needed. Users retain control over the agent's actions, with the option to interrupt, take over the browser, or stop tasks at any point.

The new agent is initially available to Pro, Plus, and Team subscribers, with access to 400 agent prompts per month for Pro users and 40 prompts for other tiers. Enterprise and Education subscribers are expected to gain access later in the summer. The feature is not yet available for free users of ChatGPT.

OpenAI's move to integrate these advanced capabilities into ChatGPT signals a strategic shift in the company's approach to AI-driven productivity tools. By offering a central hub for document and data management, analysis, and creation, ChatGPT could become a significant competitor to traditional office software like Microsoft's Excel and PowerPoint. This development aligns with OpenAI's ongoing efforts to monetize its AI models and reduce reliance on Microsoft's enterprise software.

For more information, see [1] https://www.wired.com/story/openai-chatgpt-agent-launch/, [2] https://www.testingcatalog.com/openai-readies-agent-mode-for-chatgpt-to-challenge-microsoft-office/, and [3] https://openai.com/index/introducing-chatgpt-agent/.