OpenAI Halts Launch of New GPT-6.1 Astra AI Model Citing Safety and Alignment Risks

An OpenAI logo is displayed at Moscone Center during the Dreamforce 2026 technology summit in San Francisco, California, US, on September 17, 2026.

OpenAI has officially called off the launch of its newest artificial intelligence model due to emerging safety concerns. This decision reflects a broader industry trend toward slowing down the rapid development of frontier technology.

The announcement from the AI giant on Monday follows growing anxieties regarding the potential for catastrophic harm caused by AI, especially after recent incidents involving rogue AI agents.

Saachi Jain, OpenAI’s head of safety systems, explained that the upcoming model, known as GPT-6.1 Astra, fell short of internal standards for adhering to human intentions during rigorous testing phases.

“For anything regarding safety and alignment, there’s a trade off. You really do need to find what’s the right line between staying within scope, but also avoiding laziness in terms of how the model actually pursues tasks even when it hits friction,” Jain stated in an official release provided to Al Jazeera.

Although GPT-6.1 Astra showed improvements over its predecessor in several categories, Jain noted that the model failed to reach required benchmarks concerning scope, authorization, and user communication regarding completed tasks.

“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,” Jain added. “But when we ship it to users, we have an extremely high bar in terms of safety and alignment.”

More updates to follow…