OpenAI’s new Astra model will utilize a reasoning technique known as “recurrent depth,” which allows it to function outside of the sequential thinking typical of most reasoning models, according to a report by The Information on September 2, 2026. This technique, referred to as “opaque recurrence,” may complicate the monitoring of the model’s reasoning process, raising concerns among AI safety experts.
Although Astra's application of this technique is reportedly limited, it has still generated significant apprehension among AI safety professionals. Redwood CEO Buck Shlegeris expressed his concerns in a post following the announcement, stating, “I don’t know whether Astra is much less CoT monitorable than previous models. But if OpenAI pushes this technique further, they’ll have the option to massively increase the recurrence and totally destroys CoT monitorability.”
AI safety advocate Zvi Mowshowitz also commented, suggesting that regulations may be necessary to prevent a “race to the bottom” among AI laboratories. He noted, “The technique is playing with fire, risking a taboo that OpenAI and Anthropic have fought to establish that we work hard to maintain Chain of Thought faithfulness and monitorability for as long as we can.”
Typically, a reasoning model’s chain of thought outlines the sequential steps taken while solving a problem. Despite its imperfections, this representation is valuable for monitoring potential misbehavior or misalignment. In light of OpenAI's recent issues with rogue agent activity, chain-of-thought records were instrumental in understanding the agents' behaviors.
In the case of opaque recurrence, the model adopts a less linear approach, processing the same query multiple times in a loop, which results in fewer clear traces and bypasses the conventional chain-of-thought record. However, OpenAI has indicated that Astra's use of this technique is limited, and the model's chain of thought is still expected to be understandable. The company has also refuted any claims that it would transition to “neuralese” and announced plans for comprehensive chain-of-thought monitoring systems as part of its safety initiatives.
OpenAI chief scientist Jakub Pachocki reiterated the lab’s commitment to maintaining legible chains of thought, stating, “OpenAI has worked to preserve and utilize chain-of-thought monitoring since our very first reasoning models. It’s a core goal of our current research program.”
All AI models engage in some form of opaque reasoning, and few researchers consider chain-of-thought logs as a direct representation of a model’s reasoning. Nevertheless, concerns persist that opaque recurrence could complicate AI reasoning oversight, especially as its use expands across various models. A follow-up report by The Information indicated that both Anthropic and Google DeepMind were already discussing the technique.
Ryan Greenblatt, chief scientist at Redwood Research, expressed concerns that opaque reasoning could scale more rapidly than traditional chain-of-thought reasoning, potentially obscuring all reasoning from visible channels. He stated, “My biggest concern is that a natural progression from here would involve scaling up the opaque reasoning to the point where the model reasons entirely or almost entirely in latent space. I hope it isn’t too late to avoid the most concerning architectures and that OpenAI will stop here.”