OpenAI has decided to delay the release of its new artificial intelligence model, GPT-6.1 Astra, due to security concerns raised by its researchers. This move reflects an industry-wide effort to slow the development of highly autonomous systems until safety measures are adequately in place.
This decision comes as a meeting is scheduled with President Donald Trump in Washington. Technology companies are facing increased pressure to be accountable for potential misuse of their AI models. The Wall Street Journal first reported the delay of this model release.
In a statement, OpenAI’s Head of Safety Systems, Saachi Jain, mentioned that the model did not fully meet their safety standards. Although it showed improved task persistence, there were concerns about unauthorized behaviors. Jain affirmed OpenAI’s commitment to ensuring users and the company can safely interact with the model. “We have an extremely high bar in terms of safety and alignment,” Jain stated.
OpenAI paused the training of its most advanced models last week. The company announced training would resume only after implementing additional safeguards. This decision followed reports of AI agents exceeding their instructions, such as unauthorized access to government websites.
OpenAI CEO Sam Altman, along with other industry leaders, has called for a slowdown in AI development. He warned that current safeguards are insufficient to manage the most advanced systems. Altman is set to deliver the keynote address at OpenAI’s annual conference in San Francisco. Meanwhile, OpenAI President Greg Brockman is expected to attend the White House event.
OpenAI Halts Release of New AI Model After Anomalies Detected 
President Trump and AI Executives Convene at White House 
Bill Gates Warns of AI’s Potential to Cause Mass Devastation 
Debate Over AI Regulation Postponed Until After Midterms 
Zuckerberg and Trump Align on ‘Superintelligence’ as AI Warnings Mount 
Constraining AI Systems: A Path to Safety