OpenAI rolls out new AI model GPT-6 Astra

OpenAI is rolling out GPT-6 Astra to selected customers, presenting the model as both a more powerful computer operator and a system built with stronger security protections.

The company said some cybersecurity customers would receive access on Thursday, followed by a wider release to other paying customers. Users on the free tier and the cheapest paid plan will not receive access.

OpenAI President Greg Brockman said the release reflects the risks that accompany more capable systems. “At this level of capability, safety has to become our top priority,” he told reporters on a call.

“

At this level of capability, safety has to become our top priority.

OpenAI said Astra can independently handle a broad range of time-consuming computer work, including building websites, conducting scientific analysis, developing games, working on cybersecurity and writing code. The company said an autonomous agent using the model could reduce apartment hunting from six hours to less than 10 minutes.

The model arrives after growing concern about advanced AI systems, including incidents involving products from OpenAI and Anthropic. OpenAI paused some development for two weeks during the summer after two models under testing were involved in a security breach at the AI platform Hugging Face. The company said Astra was developed with stronger safeguards and was not involved in that breach.

Brockman said, “It’s not unreasonable to feel that we are now in the AGI era,” referring to artificial general intelligence, a hypothetical stage in which AI systems match human intelligence across most tasks. OpenAI previously had an agreement with Microsoft tied to reaching AGI, but the companies scrapped the relevant exclusivity terms in April.

“

It’s not unreasonable to feel that we are now in the AGI era.