{"id":63363,"date":"2026-06-05T11:10:09","date_gmt":"2026-06-05T11:10:09","guid":{"rendered":"https:\/\/www.europesays.com\/ai\/63363\/"},"modified":"2026-06-05T11:10:09","modified_gmt":"2026-06-05T11:10:09","slug":"anthropic-says-something-unsettling-has-been-happening-to-claude","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ai\/63363\/","title":{"rendered":"Anthropic says something unsettling has been happening to Claude"},"content":{"rendered":"

\"The The logo of the US artificial intelligence safety and research company Anthropic, in Mulhouse on 21 April, 2026 (AFP via Getty Images) <\/p>\n

Leading AI firm Anthropic<\/a> has warned that humans risk losing control over AI systems in the very near future if a concerning trend with its Claude model continues.<\/p>\n

The $1 trillion startup revealed that Claude<\/a> now writes more than 80 per cent of its own code \u2013 up from less than 10 per cent in February last year.<\/p>\n

Anthropic claims that the Claude-written code is currently roughly at parity to human-written code, and is expected to be much better within the next year.<\/p>\n

\u201cThat trend points to an AI system capable of fully autonomously designing and developing its own successor,\u201d Anthropic\u2019s Marina Favaro and Jack Clark wrote in a blog post<\/a> on Thursday.<\/p>\n

\u201cAI that can build itself would be a major development in the history of technology \u2013 one that could bring enormous good for the world in science, healthcare, and beyond.<\/p>\n

\u201cBut full recursive self-improvement also might increase the risks of humans losing control over AI systems.\u201d<\/p>\n

Anthropic noted that, until recently, artificial intelligence<\/a> advances have relied entirely on human engineers and developers. But the company is now increasingly delegating the development of its AI to the AI systems themselves in an effort to speed up progress.<\/p>\n

These advances are putting Claude on a path towards \u201crecursive self-improvement\u201d, which could see it improve itself without human intervention.<\/p>\n

Anthropic added that it would \u201clikely be a good thing\u201d to effectively slow the development of the technology in order to give more time to deal with the potential implications.<\/p>\n

A slowdown or pause of AI development would require broad industry consensus, as well as political cooperation amid intense geopolitical pressures and rivalries.<\/p>\n

\u201cA meaningful slowdown or pause would require multiple well-resourced labs at or near the frontier, in multiple countries, agreeing to stop under the same conditions,\u201d Anthropic\u2019s post stated.<\/p>\n

\u201cIt would also require that each can verify that the others have actually stopped… None of this is necessarily impossible in principle \u2013 the world has built verification regimes for other complex technologies \u2013 but those regimes took decades to build both the infrastructure and the trust. We don\u2019t have that long.\u201d<\/p>\n

Anthropic said it plans to work with policymakers, researchers and other AI companies in the coming months in an effort to prepare for the arrival of self-improving AI systems.<\/p>\n

It is not the first time there has been such a call to pause AI development, with thousands of technologists calling for a six month moratorium on AI training in 2023 following the release of ChatGPT.<\/p>\n

The open letter, organised by the non-profit Future of Life Institutes, cited the danger of society-wide loss of control of AI, though no pause was achieved.<\/p>\n","protected":false},"excerpt":{"rendered":"The logo of the US artificial intelligence safety and research company Anthropic, in Mulhouse on 21 April, 2026…\n","protected":false},"author":2,"featured_media":63364,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[8],"tags":[53,3154,182,633,36087],"class_list":["post-63363","post","type-post","status-publish","format-standard","has-post-thumbnail","category-anthropic","tag-anthropic","tag-anthropic-claude","tag-claude","tag-development","tag-losing-control"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/63363","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=63363"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/posts\/63363\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media\/63364"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/media?parent=63363"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/categories?post=63363"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ai\/wp-json\/wp\/v2\/tags?post=63363"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}