OpenAI has introduced GPT-6 Astra, a new artificial intelligence model designed to operate computers and complete complex tasks on behalf of users, marking another major development in the push toward more autonomous AI systems.
The launch comes as competition between OpenAI and rival AI firm Anthropic continues to intensify, accelerating the development of increasingly capable artificial intelligence technologies.
Astra Demonstrates Advanced Computer Skills
In a demonstration, Astra was shown carrying out a range of tasks across different computer applications. These included turning a basic drawing into a detailed rocket window, creating a 3D model using Blender, designing a presentation and preparing an online listing for a second-hand table.
The model was also demonstrated building a computer game, drafting a licensing agreement, finding a tennis court and preparing a file for a 3D printer.
The capabilities demonstrate a broader shift in AI technology—from systems primarily designed to answer questions to tools that can independently perform tasks and take actions for users.
OpenAI CEO Sam Altman encouraged people to challenge the model with increasingly complex assignments.
He said: “What I hope will happen is people, I think people will be surprised at the beginning, but then as they build up more trust in the model… just start throwing harder and harder tasks and more creative ideas at it.”
Altman said AI can already handle a wide range of activities, including developing computer games, running scientific simulations, financial modelling, creating presentations and building interactive software.
Growing AI Power Raises Security Concerns
As AI systems become more capable, concerns over cybersecurity are also increasing.
OpenAI has limited some cybersecurity capabilities in its latest models amid concerns that advanced AI could potentially be used to identify vulnerabilities or gain unauthorised access to computer systems.
The rapid development of generative AI has also renewed discussions about how children should interact with the technology.
New York City Mayor Zohran Mamdani announced a one-year moratorium on generative AI for students from 2-K through eighth grade, arguing that young people require teachers and human interaction as part of the learning process.
“Children need their teachers and human connection in order to learn and in order to grow,” he said.
Officials Demand Greater Public Communication
US Treasury Secretary Scott Bessent also criticised AI companies over their communication with the public, arguing that the industry has not done enough to explain the technology and its potential economic impact.
He said AI companies had done a “horrendous job” explaining the technology and assuring the public that its economic benefits would be widely shared.
Altman acknowledged that AI companies could have communicated more effectively but maintained that society must have open discussions about the technology’s potential risks and work collectively to address them.
With Astra demonstrating an increasing ability to browse the internet, create content, write code and execute complex tasks, the AI industry is moving toward a new phase—one where artificial intelligence is not simply answering questions but actively getting work done.
