Editorial illustration for OpenAI launches GPT-5.4 and ChatGPT Agent, enabling computer‑task automation
GPT-5.4 Automates Desktop Tasks with ChatGPT Agent
OpenAI launches GPT-5.4 and ChatGPT Agent, enabling computer‑task automation
OpenAI has stopped asking for permission. With GPT-5.4, the company isn't just releasing another text generator. It's deploying a system that takes over your computer.
That shift is embodied in the new ChatGPT Agent, a tool designed to click, type, and browse autonomously. Its method is direct: the AI writes executable code to manipulate your keyboard and mouse, guided by screenshots it captures itself. The stated goal, like shopping for groceries online, is a system that completes multi-step digital tasks alone.
This isn't an incremental update. It's a decisive leap from an assistant that talks to one that acts.
OpenAI introduced ChatGPT Agent amid a flurry of other agentic tools that emerged last year, which can take control of your computer to perform tasks, such as searching for and buying ingredients for a meal. While OpenAI is bringing GPT-5.4 to its API and its AI-powered coding tool, Codex, it's rolling out its reasoning model, GPT-5.4 Thinking, to ChatGPT. OpenAI says GPT-5.4 can write code to operate computers, as well as issue keyboard and mouse commands in response to screenshots. GPT-5.4 also shows improvements while using web browsers, as well as its ability to call upon tools and APIs more accurately and efficiently to help it complete tasks.
The technical leap is real. It fundamentally changes the risk. The problem, as The Verge outlines, is no longer about crafting a decent email.
It's about trust. Can you trust this agent with your passwords? Your credit card?
The autonomy of your own desktop? The practical consequence is a profound new dependency. We are building the ghost in the machine.
Now we're handing it the keys.
Common Questions Answered
How does OpenAI's GPT-5.4 enable computer-task automation?
GPT-5.4 can write code to operate computers and issue keyboard and mouse commands in response to screen prompts. The model can navigate between multiple applications like spreadsheets, presentations, and web browsers, performing tasks autonomously with user supervision.
What capabilities does the new ChatGPT Agent introduce?
The ChatGPT Agent can take control of a computer to perform complex tasks, such as searching for and purchasing ingredients for a meal. It represents a significant advancement in AI's ability to interact with computer interfaces and execute multi-step instructions.
How does GPT-5.4 differ from previous OpenAI language models?
GPT-5.4 introduces an enhanced reasoning model called GPT-5.4 Thinking, which extends beyond traditional language processing to include coding and computer operation capabilities. The model blends reasoning, coding, and document work, offering a more versatile approach to task automation.
Further Reading
- GPT-5.4 Leaked: 2M Context Window, Full-Resolution Vision, and Enhanced Agent Capabilities — NxCode
- Introducing ChatGPT agent: bridging research and action — OpenAI Official Blog
- OpenAI launches a way for enterprises to build and manage AI agents — TechCrunch
- OpenAI GPT-5 Model Guide: GPT-5.2 vs 5.3 vs 5.4 — NxCode
- GPT-5.4 LEAKED: Early Look at the New Features! — Universe of AI (YouTube)