{"id":10022,"date":"2026-10-16T08:41:16","date_gmt":"2026-10-16T08:41:16","guid":{"rendered":"https:\/\/cybersecurityinfocus.com\/?p=10022"},"modified":"2026-10-16T08:41:16","modified_gmt":"2026-10-16T08:41:16","slug":"cyberattack-ais-must-be-trained-only-on-air-gapped-systems-hundreds-of-ai-researchers-say","status":"publish","type":"post","link":"https:\/\/cybersecurityinfocus.com\/?p=10022","title":{"rendered":"Cyberattack AIs must be trained only on air-gapped systems, hundreds of AI researchers say"},"content":{"rendered":"<div>\n<div class=\"grid grid--cols-10@md grid--cols-8@lg article-column \">\n<div class=\"col-12 col-10@md col-6@lg col-start-3@lg\">\n<div class=\"article-column__content\">\n<div class=\"container\"><\/div>\n<p class=\"wp-block-paragraph\">Current security procedures for training AI models are inadequate, hundreds of AI researchers have warned governments and the AI community. They are calling for tougher rules and the establishment of a new international body overseeing best practice for AI development.<\/p>\n<p class=\"wp-block-paragraph\">In an <a href=\"https:\/\/ai-security-letter.org\/\" target=\"_blank\" rel=\"noopener\">open letter<\/a> the AI specialists \u2014 487, at time of publication, including many senior academics \u2014 called for more AI safety research and greater emphasis on containment technologies.<\/p>\n<p class=\"wp-block-paragraph\">\u201cStrong sandboxing mechanisms including air-gapping (disconnecting from the internet) should be used to provide high-assurance containment of cyber-capable AI models,\u201d say the signatories, who include such luminaries as Geoffrey Hinton and Yoshua Bengio.<\/p>\n<p class=\"wp-block-paragraph\">The experts said that a new international body, to be set up in consultation with governments and a wide range of AI and security specialists, could set new safety standards but also act as an organization to which containment failures could be reported, improving transparency for learning and accountability purposes.<\/p>\n<p class=\"wp-block-paragraph\">The scientists\u2019 demands go further than some recent calls for greater regulation of AI. Earlier this week, Microsoft president Brad Smith said <a href=\"https:\/\/www.bbc.com\/news\/articles\/c6rmy7zmvx3lo\" target=\"_blank\" rel=\"noopener\">AI is less regulated than washing machines<\/a> or milk, although he made no firm proposals to rectify the situation. And the <a href=\"https:\/\/committees.parliament.uk\/committee\/365\/business-innovation-science-and-trade-committee\/news\/218156\/house-of-commons-committee-opens-major-new-inquiry-into-ai-security-risks\/\" target=\"_blank\" rel=\"noopener\">UK parliament is looking at ways to improve AI security<\/a>, although any moves will be hampered by disagreement among AI businesses.\u00a0<\/p>\n<p class=\"wp-block-paragraph\">Hinton and Bengio are no strangers to this kind of scientist-led initiative. In October last year, they signed an open letter <a href=\"https:\/\/www.computerworld.com\/article\/4077074\/call-to-ban-ai-superintelligence-could-redraw-the-global-tech-race-between-the-us-and-china.html\">warning of the dangers of developing super-intelligent AI<\/a>. The letter appears not to have affected AI\u2019s breakneck development; it remains to be seen whether AI players in boardrooms will heed the warning.<\/p>\n<p class=\"wp-block-paragraph\">\n<\/p><\/div>\n<\/div>\n<\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>Current security procedures for training AI models are inadequate, hundreds of AI researchers have warned governments and the AI community. They are calling for tougher rules and the establishment of a new international body overseeing best practice for AI development. In an open letter the AI specialists \u2014 487, at time of publication, including many [&hellip;]<\/p>\n","protected":false},"author":0,"featured_media":10023,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[],"class_list":["post-10022","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-education"],"_links":{"self":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts\/10022"}],"collection":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=10022"}],"version-history":[{"count":0,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/posts\/10022\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=\/wp\/v2\/media\/10023"}],"wp:attachment":[{"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=10022"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=10022"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cybersecurityinfocus.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=10022"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}