{"id":77397,"date":"2026-06-17T20:00:14","date_gmt":"2026-06-17T20:00:14","guid":{"rendered":"https:\/\/www.europesays.com\/ai\/77397\/"},"modified":"2026-06-17T20:00:14","modified_gmt":"2026-06-17T20:00:14","slug":"catholic-expert-commends-anthropic-for-global-call-to-slow-down-ai-check-values","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ai\/77397\/","title":{"rendered":"Catholic expert commends Anthropic for global call to slow down AI, check values"},"content":{"rendered":"

\"\"<\/a>Anthropic’s logo is seen in this illustration taken March 1, 2026. In a June, 4, 2026, blog post, Anthropic officials urged a slowdown in AI development, warning of the risks of humans “losing control over AI systems” if they become “fully capable” of building their own successors. (OSV News illustration\/Dado Ruvic, Reuters)
\n <\/a><\/p>\n

A Catholic expert on artificial intelligence is commending recent calls by AI research firm Anthropic to rein in the technology amid a significant acceleration in its growth.<\/p>\n

Anthropic \u2014 which recently joined Pope Leo XIV in unveiling his new encyclical on AI<\/a> \u2014 broadly called for \u201ca meaningful slowdown or pause\u201d in advanced AI development, as the technology is poised to accelerate \u201csooner than most institutions are prepared for.\u201d<\/p>\n

That message was issued in a June 4 post on the Anthropic Institute website, with authors Marina Favaro and Jack Clark explaining the concept of \u201crecursive self-improvement\u201d \u2014 or, in the words of their article\u2019s title, \u201cWhen AI builds itself.\u201d<\/p>\n

<\/a><\/p>\n

Through recursive self-improvement, an AI system uses existing AI technology to improve upon itself, potentially becoming \u201ccapable of fully autonomously designing and developing its successor,\u201d Favaro and Clark said.<\/p>\n

Recursive self-improvement is \u201ceffectively the creation of an automated AI engineer, or many, many, many automated AI engineers (as many as they have the computational capacity to run),\u201d explained Brian Patrick Green, director of technology ethics at the Markkula Center for Applied Ethics at Santa Clara University. \u201cRather than humans programming AI, AI becomes powerful enough to program itself.\u201d<\/p>\n

Green added that such self-improvement is \u201cone of the goals for AGI,\u201d or \u201cartificial general intelligence.\u201d<\/p>\n

But, warned Favaro and Clark, while self-building AI could bring either \u201cenormous good for the world\u201d \u2014 for example, in science and healthcare \u2014 \u201cfull recursive self-improvement also might increase the risks of humans losing control over AI systems.\u201d<\/p>\n

Others hold that the potential loss of human oversight is more nuanced, with humans retaining varying levels of control amid rapid AI growth, as author Mathew Hutson noted in a May 7 article for IEEE Spectrum, the flagship publication of the global technical professional organization IEEE, the Institute of Electrical and Electronics Engineers.<\/p>\n

Still, said Green, Anthropic\u2019s concerns are well-founded, especially given the critical role of alignment, or the process of ensuring the technology squares with human values, so that AI models safely serve human interests.<\/p>\n

\u201cRecursive self-improvement can lead to a situation called \u2018values lock-in,’\u201d said Green.<\/p>\n

In such a scenario, he said, \u201cwhatever the values are at the start of improvement become locked-in for all subsequent models because the machine is really in charge of making new versions of itself.\u201d<\/p>\n

That includes \u201cchoosing its values,\u201d Green said, noting that \u201chumans might no longer be in a position to control the machine\u2019s behavior.\u201d<\/p>\n

\u201cIt is very important to get recursive self-improvement right, otherwise it could spiral out of control and create new versions of itself that are not friendly,\u201d he said.<\/p>\n

Favaro and Clark underscored that any downshifting in frontier AI development would require a \u201cglobal coordination mechanism\u201d in order to prevent bad actors from exploiting such a pause.<\/p>\n

Green also commended Anthropic\u2019s move to comply (albeit grudgingly) with a June 12 Trump administration order to suspend foreign access to the company\u2019s Fable 5 and Mythos 5 models due to national security concerns.<\/p>\n

The two AI models have, among other features, strong performance levels in agentic coding (displaying reasoning and programming abilities), knowledge work (providing expertise and critical thinking) and spatial reasoning (representing objects in relation to each other).<\/p>\n

In a statement that same day, Anthropic said it would remove access to the two AI models \u201cfor all users,\u201d stressing, \u201cWe disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people.\u201d<\/p>\n

The company reiterated its earlier statement that \u201cgovernment should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts.\u201d<\/p>\n

However, said Anthropic<\/a>, \u201cthis action does not adhere to those principles.\u201d<\/p>\n

The company regarded the Trump directive as \u201ca misunderstanding,\u201d and planned to \u201crestore access as soon as possible.\u201d<\/p>\n

Regarding the federal halt, Green said it was \u201cbetter to be safe than sorry.\u201d<\/p>\n

\u201cMythos has demonstrated powerful cyberoffense capabilities, and it is understandable that the U.S. government would not want enemy nations to gain access to such a tool,\u201d he said.<\/p>\n

He noted that \u201cAnthropic itself has recommended this caution, so the disagreement seems to be whether the restrictions are good enough, which is a matter of judgment.\u201d<\/p>\n

Green admitted that \u201cin some ways, the restrictions might be too stringent in that even Anthropic\u2019s own non-U.S. citizen employees \u2014 many of whom are from friendly nations like Canada, the U.K. and Germany \u2014 are no longer allowed access.\u201d<\/p>\n

That move signals \u201cdirect government interference with business activity, and should cause American businesses some concern about government overreach,\u201d said Green.<\/p>\n

<\/a> <\/p>\n","protected":false},"excerpt":{"rendered":"Anthropic’s logo is seen in this illustration taken March 1, 2026. In a June, 4, 2026, blog post,…\n","protected":false},"author":2,"featured_media":77398,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[8],"tags":[24,53,25,4977,24247],"class_list":["post-77397","post","type-post","status-publish","format-standard","has-post-thumbnail","category-anthropic","tag-ai","tag-anthropic","tag-artificial-intelligence","tag-artificial-intelligence-ethics","tag-encyclical"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/77397","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/comments?post=77397"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/77397\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/77398"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=77397"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=77397"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=77397"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}