GPT-6 Astra System Card - OpenAI Deployment Safety Hub

In this section, we provide results from our investigation into Astra’s monitorability, that is, the extent to which we can expect monitoring systems to be able to detect when Astra acts in misaligned or otherwise undesirable ways. According to our evaluations, GPT-6 Astra shows a substantial decrease in chain-of-thought monitorability compared to previous models. Overall our alignment evaluations show that Astra is more likely than GPT-5.6 Sol to respect security and safety restrictions which make us confident in still deploying this model to the wider public.

We are actively investigating the causes of GPT-6 Astra’s increased CoT controllability. Unlike previous models , we find that CoT controllability for Astra especially increases over the course of RL training. We are fairly confident that there are no significant direct CoT grading pressures. We are also quite confident that changes in CoT controllability are not differentially due to any architectural changes. We have found that these changes are correlated with the increase in no-CoT capabilities (i.e., being able to take actions or return an answer without any verbalized reasoning required)

Overall, most of the aggregate gap in CoT monitorability comes from GPT-6 Astra producing shorter CoTs that often omit or weaken the evidence the monitor needs, including by producing empty or nearly empty CoTs more often. In ShadeArena, Impossible Coding Tasks, and Health Queries: Missing Evidence, those gaps largely close at longer matched CoT lengths. Meanwhile, on Destructive Actions and Health Queries: Patient Opinion, GPT-6 Astra continues to have lower monitorability even at longer matched CoT lengths relative to GPT-5.6 Sol. In these instances, the intervention shifts GPT-6 Astra’s answers distributionally, but the CoT gives an independent-looking rationale for the answer, making the intervention’s influence hard to detect.