OpenAI has canceled plans to release GPT-6.1 Astra, its next AI model, after the system failed to meet the company’s internal safety standards during testing. The decision comes as Washington prepares for a high-profile gathering on artificial intelligence policy.
Saachi Jain, OpenAI’s head of safety systems, said the company applies an “extremely high bar” before putting any model in front of the public.
“Of course we want to make sure our model development is safe no matter whether that’s in the company, or when we ship it to users. But when we ship it to users, we have an extremely high bar in terms of safety and alignment.”
OpenAI released GPT-6 Astra earlier this month, but the follow-up version will not be joining it on the market. The announcement landed a day before the company’s annual developers conference. Separately, OpenAI CEO Sam Altman has backed a call from Anthropic for the AI industry to slow its pace of development rather than race to release increasingly capable systems.
Trump’s Golden Age Event
President Donald Trump is set to headline the “Golden Age” event Tuesday at the Andrew W. Mellon Auditorium in Washington, D.C., with artificial intelligence, energy and space on the agenda. Attendees listed on the invitation include Elon Musk, Nvidia CEO Jensen Huang, Anthropic co-founder Tom Brown and Palantir Chief Technology Officer Shyam Sankar.
Trump has resisted calls to slow American AI development, citing concerns that doing so would benefit China. Huang has also questioned some warnings from safety advocates. Even so, OpenAI and Anthropic have taken those concerns seriously enough to discuss slowing development publicly — and in Astra’s case, to halt a release altogether.
The Trump administration has set up a voluntary process allowing companies to submit powerful AI models to the federal government for review before public release, while the president continues to warn against regulations that could put American developers at a disadvantage against China.
A Vermont Senator’s Warning
The safety debate has also reached Capitol Hill. Sen. Bernie Sanders (I-VT) called senators to a private briefing on what he described as the “extraordinary dangers that AI poses for humanity,” citing warnings from Jacob Coxon, a former Anthropic researcher who resigned after roughly four months at the company. Coxon has said AI companies are “gambling with our lives” and that increasingly powerful systems could eventually threaten humanity. Sanders has also backed legislation that would pause AI development.
A professional communications firm with ties to the AI-safety movement was reported to be working with Coxon as his warning spread through the media, even though Coxon told Fox News’ Bret Baier that he had not worked with third parties on his public rollout.
The Astra decision gives policymakers a concrete example to weigh against those warnings: OpenAI tested a model, determined it did not clear the company’s safety threshold, and withheld it from the public rather than release it.
Questions remain about what happens if a company’s judgment on safety proves wrong, or if standards differ across firms. But in this instance, OpenAI chose not to ship a product its own safety team would not approve — a decision now sitting alongside broader debates in Washington over how fast, and how cautiously, the AI industry should move.
You must be logged in to post a comment Login