ChatGPT can now do work for you using its own computer.

7,769,590 followers

2w Edited

ChatGPT can now do work for you using its own computer. Introducing ChatGPT agent—a unified agentic system combining Operator’s action-taking remote browser, deep research’s web synthesis, and ChatGPT’s conversational strengths. Built-in safety keeps you in control: ✅ Confirmation before important actions ✅ No data stored in takeover mode ✅ One-click to clear browsing and log out of all sites ChatGPT agent starts rolling out today to Pro, Plus, and Team users. Pro users will get access by the end of day, while Plus and Team users will get access over the next few days. Enterprise and Edu users will get access in the coming weeks. https://lnkd.in/esn2WSEV

624 Comments

Urica Lopez

🤖 AI-Driven Workflow Architect | Project & Ops Manager | Branding Strategist | Founder & Creative Thinker

Wow, this is exciting! 🤩 ChatGPT is basically the IQ points and extra hands I wasn’t born with, it completes me. 🤖🧠

2 Reactions

Alessandro Pregnolato

I make companies data-driven by building scalable data teams, infrastructures, frameworks, and processes. Data Advisor/Consultant for VCs & Scale-ups | Fractional VP Data | ex Preply, Typeform, King, Adobe.

OMG such a stream of generic enthusiastic comments (Great! Amazing! Game changer!!). Is there any critical sense left at all in this world. I can't hardly think of any practical applications.

2 Reactions

Luv Chanan (Oliver)

AI is cool

This is a massive leap toward autonomous AI — but also a new attack surface we need to prepare for. A browser-capable agent with task execution opens powerful possibilities — but also raises critical questions: • Can agents detect and avoid deceptive websites, phishing pages, or malicious redirects? • How will session hijacking or persistent cookies be mitigated across multiple runs? • Could a hijacked prompt chain turn an agent into a weaponized automation bot? Huge kudos to OpenAI for prioritizing: ✅ Confirmation before actions ✅ No data persistence in takeover mode ✅ One-click memory clearing These guardrails are essential. As agentic systems evolve, secure autonomy may become the next frontier in AI safety — and privacy-first design will be non-negotiable.