OpenAI begins rollout of new powerful AI model GPT-6 Astra

OpenAI has announced the upcoming release of its latest and most advanced AI model, GPT-6, also known as Astra, to select customers. President Greg Brockman emphasized that safety is now the top priority with this level of capability during a call about the release. Just over a year has passed since the previous version, GPT-5, was launched.

Concerns about advanced models' capabilities have surged since then, following incidents involving systems developed by OpenAI and its rival, Anthropic. OpenAI temporarily halted some model development for two weeks this summer after two models tested were involved in a security breach at AI platform Hugging Face. Although the breached models were not Astra, the company has implemented stronger safeguards in response.

The new Astra model will be accessible to certain cybersecurity customers starting Thursday, with a broader rollout to other paying customers planned. Free tier and cheapest paid plan users will not have access. OpenAI claims Astra can autonomously tackle a wide array of tedious computer tasks, such as website creation, scientific analysis, game development, cybersecurity, and coding.

To demonstrate the time-saving potential, the company stated that Astra could reduce apartment hunting from six hours to under 10 minutes. Brockman suggested that we might now be in the era of artificial general intelligence (AGI), an advanced stage where AI systems match human intelligence across most tasks. The previous agreement with Microsoft, a significant investor, would end once OpenAI reached AGI, but those terms were abandoned in April.