OpenAI confirmed Monday that it would not release the newest AI model, called GPT-6.1 Astra, after it “didn’t quite meet the bar” for safety.
The company’s confirmation came after the Wall Street Journal reported earlier on Monday that the release of GPT-6.1 Astra had been scrapped ahead of its planned October debut.
OpenAI previously described the Astra model as “state-of-the-art on computer use, browsing, professional work, software engineering, cybersecurity and science.” However, in a statement to media, Saachi Jain, OpenAI’s head of safety systems, explained that the new model showed what the company viewed as high levels of deception. Additionally, it’s said the model was willing to go beyond what it was originally asked to do, notably not checking back for further directions.
“For anything regarding safety and alignment, there’s a trade-off,” Jain told media Monday. “[Astra] didn’t quite meet the bar in terms of staying within scope and authorization, and how it communicates back to the user about the type of work it’s done.”
As Jain went on, she noted that there’s an “extremely high bar” for their models, in terms of safety expectations, before it can be released to consumers. She added: “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.”
The decision to shelve GPT-6.1 Astra comes on the heels of several reports claiming rogue behavior from OpenAI’s models amid testing, including hacking websites without the company knowing, as well as meddling with U.S. government websites.
“There is an extensive and ongoing review related to our agents’ use of internet access during training and evaluation. We’ve been publishing summaries at the link below and will continue to,” OpenAI CEO Sam Altman said on X Friday. “We have not been as fast as we would have liked but we are trying to balance our desire for transparency with gaining a clear understanding from petabytes of agent activity logs, and working with impacted organizations.”
It also follows Anthropic CEO Dario Amodei calling for a “pacing the frontier” amid AI safety concerns. Several tech leaders co-signed this sentiment, including Altman and Elon Musk, but was met with derision from President Donald Trump.
In response, Trump wrote on Truth Social on Sept. 14, “The only control or ‘guardrails’ that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades! The Trump Administration has stopped AI ‘people’ from doing bad, or potentially bad, ‘things,’ like Dario (Anthropic!), who is now pretending to be a ‘perfect little angel’ – and we will continue to do so!”
Though, Trump and Amodei have since met to discuss the issue.
