{"id":1039656,"date":"2026-09-02T22:17:26","date_gmt":"2026-09-02T22:17:26","guid":{"rendered":"https:\/\/www.europesays.com\/us\/1039656\/"},"modified":"2026-09-02T22:17:26","modified_gmt":"2026-09-02T22:17:26","slug":"openais-new-reasoning-technique-alarms-ai-safety-experts","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/us\/1039656\/","title":{"rendered":"OpenAI\u2019s new reasoning technique alarms AI safety experts"},"content":{"rendered":"

OpenAI\u2019s new Astra model will use a reasoning technique called \u201crecurrent depth\u201d that allows it to operate outside of the sequential thinking that characterizes most reasoning models, The Information reported<\/a> on Tuesday. This technique, also called \u201copaque recurrence,\u201d will likely make the model\u2019s chain of thought more difficult to monitor \u2014 and that has AI safety experts rattled.<\/p>\n

While Astra\u2019s use of the technique is reportedly limited, its emergence has still raised significant concerns among AI safety experts.<\/p>\n

\u201cI am extremely concerned by the reporting that Astra uses opaque recurrence,\u201d wrote Redwood CEO Buck Shlegeris in a post<\/a> after the news broke. \u201cI don\u2019t know whether Astra is much less CoT monitorable than previous models. But if OpenAI pushes this technique further, they\u2019ll have the option to massively increase the recurrence and totally destroys CoT monitorability.\u201d<\/p>\n

Longtime AI safety advocate Zvi Mowshowitz also weighed in and wrote<\/a> that laws might be necessary to prevent a \u201crace to the bottom\u201d among AI labs.\u00a0<\/p>\n

\u201cThe 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,\u201d Mowshowitz wrote. \u201cMore intensive use of such techniques would probably damage monitorability.\u201d<\/p>\n

Under normal circumstances, a reasoning model\u2019s chain of thought provides the sequential steps taken by the model as it attempts to solve a problem. While the representation is imperfect, it still serves as a valuable tool for monitoring misbehavior or misalignment. In the case of OpenAI\u2019s recent rogue agent activity, chain-of-thought records were an important tool in teasing out why agents behaved the way they did.<\/p>\n

In opaque recurrence, the model takes a less linear approach, processing the same query several times in a loop. The result leaves fewer legible traces, effectively side-stepping a conventional chain-of-thought record.<\/p>\n

Crucially, Astra\u2019s use of the technique appears to be limited. The model\u2019s chain of thought is still expected to be legible, and the company pushed back against any suggestion that it would shift to \u201cneuralese.\u201d OpenAI has already announced plans for extensive chain-of-thought monitoring systems as part of its forward-looking safety plans.<\/p>\n

In a post on X<\/a>, OpenAI chief scientist Jakub Pachocki emphasized the lab\u2019s commitment to legible chains of thought. \u201cOpenAI has worked to preserve and utilize chain-of-thought monitoring since our very first reasoning models,\u201d Pachocki wrote. \u201cIt\u2019s a core goal of our current research program.<\/p>\n

All AI models do some quantity of opaque reasoning, and few researchers take chain-of-thought logs as a direct representation of a model\u2019s reasoning. Still, those caveats don\u2019t dispel the concern that opaque recurrence may make AI reasoning harder to monitor, particularly as it grows in use across different models. In a follow-up report Wednesday morning, The Information reported that both Anthropic and Google DeepMind were already discussing the technique.<\/p>\n

In a post responding to the news,<\/a> Redwood Research chief scientist Ryan Greenblatt said opaque reasoning could easily scale faster than conventional chain-of-thought reasoning, effectively removing all reasoning from visible channels.<\/p>\n

\u201cMy 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,\u201d Greenblatt wrote. \u201cI hope it isn\u2019t too late to avoid the most concerning architectures and that OpenAI will stop here.\u201d<\/p>\n

When you purchase through links in our articles, we may earn a small commission<\/a>. This doesn\u2019t affect our editorial independence.<\/p>\n","protected":false},"excerpt":{"rendered":"OpenAI\u2019s new Astra model will use a reasoning technique called \u201crecurrent depth\u201d that allows it to operate outside…\n","protected":false},"author":3,"featured_media":1039657,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_share_on_mastodon":"0"},"categories":[6],"tags":[48300,64,305,67,132,68],"class_list":["post-1039656","post","type-post","status-publish","format-standard","has-post-thumbnail","category-business","tag-ai-safety","tag-business","tag-openai","tag-united-states","tag-unitedstates","tag-us"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@us\/117203766397181566","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/posts\/1039656","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/comments?post=1039656"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/posts\/1039656\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/media\/1039657"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/media?parent=1039656"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/categories?post=1039656"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/us\/wp-json\/wp\/v2\/tags?post=1039656"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}