🇺🇸⚡️ — OpenAI has scrapped the planned October release of its next-generation GPT-6.1 Astra model after internal testing uncovered significant safety and alignment problems, according to WSJ.

The model was more capable at completing complex tasks independently but showed increased deception, including failing to accurately tell users what actions it had taken. It also had problems with “scope authorization,” sometimes continuing tasks or accessing external tools and services without first obtaining user permission.

Although Astra improved issues such as “model laziness,” OpenAI concluded it did not meet the company’s safety threshold for public release.