SAN FRANCISCO, September 3 — OpenAI unveiled GPT-6 Astra on Thursday, a new artificial intelligence model the San Francisco company describes as a major step forward in computer use, coding, scientific research and professional work. The model began rolling out Thursday to enterprise customers with Daybreak access. OpenAI says Astra will reach ChatGPT Plus, Pro, Business and Enterprise users in the coming days, and will also be available through the OpenAI API and Amazon Web Services.
Here’s what it means in plain terms. Astra is built on advances across pre-training, reinforcement learning and alignment, and OpenAI calls it state-of-the-art on computer use, browser use, software engineering, cybersecurity, science and professional work.
Think of it like hiring an assistant who doesn’t just answer questions but actually sits down at the keyboard and does the job — filling out online forms, updating customer records, organizing calendars and running research. It can also produce documents, spreadsheets and presentations.
A new bar for software and science
OpenAI calls GPT-6 Astra “the best model for software engineering to date” and a “major advance for scientific discovery, mathematics, and health.” The company adds that Astra “excels at exercising care, respecting task boundaries, and communicating transparently.” CEO Sam Altman, speaking to FOX Business Network’s “The Claman Countdown” after the launch, called Astra the company’s “most aligned model ever” and said it sets “a new standard” for safety as AI models grow more capable. “We have set a new standard with this model.
It’s why it took us a while to get it out, but we think it’ll be worth the wait,” Altman said. One capability Altman singled out was Astra’s habit of spotting problems users hadn’t considered.
He described a complex chip supply chain research task in which the model flagged parts of the supply chain the user hadn’t thought to ask about.
Brockman and the AGI question
OpenAI President Greg Brockman, speaking to reporters before the launch, framed Astra as a turning point. “I think it’s not unreasonable to feel that we are now in the AGI era,” Brockman said. OpenAI defines artificial general intelligence as “highly autonomous systems that outperform humans at most economically valuable work,” according to its website.
On cybersecurity, OpenAI says Astra “is a significant jump in cyber capabilities and meets the Critical threshold in cybersecurity under our Preparedness Framework,” referencing its safety update. The launch lands against a busy backdrop for AI safety news.
OpenAI previously said an AI model hacked another company’s systems during an internal test, and the company joined more than 100 others in warning about a coming surge in AI-powered cyberattacks and calling for a global defense push. Broader industry developments include Anthropic AI model warnings about a 12% unemployment rate despite a massive GDP boost, an Anthropic researcher quitting with a grim warning over the AI race, and ongoing commentary on AI data centers and workforce productivity.
What to watch next is how quickly Astra reaches everyday ChatGPT users and whether the safety claims hold up under outside scrutiny. For now, OpenAI is betting that a model which does the tedious work — and flags what you missed — is the one people will actually keep using.
