OpenAI Begins GPT-6 Astra Rollout After Warning of Critical Cyber Capabilities

In an announcement published on its website, OpenAI said Astra represents “a significant step up in cyber capabilities” and meets the company’s “Critical” threshold. Under the right circumstances, OpenAI said, the tool can find “previously unknown security flaws and develop new ways to exploit them across many well-protected systems without a person guiding each step.” OpenAI also said Astra improves computer use, software engineering and science, and is better at staying oriented than previous versions.

Roughly a month ago, OpenAI said those capabilities prompted it to pause some internal development and trigger safety protocols. The company scaled up certain security controls and paused activities that did not meet its newly established security requirements. It then decided Astra was ready for a staggered release, but not for the general public, starting with a limited group of companies in the Daybreak program, a cybersecurity initiative that gives vetted organizations specialized AI tools for vulnerability research and threat detection.

The rollout frustrated many ChatGPT users, especially paying subscribers. TechRadar reported that all Plus, Pro, Business and Enterprise users were initially supposed to get Astra, along with customers using OpenAI API, Microsoft Azure and AWS Bedrock. Hours after launch, OpenAI chief executive Sam Altman apologized for what The Verge described as a “messy rollout.” In a post on X, Altman said: “We are working towards getting Astra in everyone’s hands as quickly as we can. I know it is frustrating and I appreciate the patience. It should be quick.” CNBC reported the exchange.

Editor’s Summary: OpenAI has begun a phased release of GPT-6 Astra after warning that the model can autonomously find and exploit unknown security flaws. Initial access is restricted to vetted Daybreak program participants, and OpenAI’s chief executive apologized for the delays while saying wider access should be quick.