Cybersecurity experts are calling it \u201cBlack October.\u201d\u00a0<\/p>\n
Earlier this week, researchers disclosed two major vulnerabilities<\/a> hidden in PHP, a programming language upon which roughly seven out of ten websites are built. Since then, developers have been racing to patch their systems before they\u2019re hijacked by malicious hackers. Websites around the world intentionally went offline<\/a> as businesses and governments worked to implement the fix. OpenAI announced it would spearhead a global effort<\/a> to find similar vulnerabilities still awaiting discovery. Some believe it\u2019s just the initial tremor of a much more serious cyber earthquake to come: \u201cas bad as things have gotten in the past 4 days, they are going to get a lot worse before they get better,\u201d Anthropic cybersecurity lead Jason Clinton said in a Thursday X post<\/a>. (\u201cIt was always going to be this way. We tried to warn the world,\u201d he said in another post<\/a> responding to a claim that the AI companies themselves had been responsible for the upheaval.)<\/p>\n Amid this all-hands-on-deck cybersecurity panic, AI industry enfant terrible Abliteration doesn\u2019t appear to be all that concerned.<\/p>\n The Palo Alto-based tech company, building what it describes<\/a> as \u201cunrestricted AI,\u201d has just launched a cheaper (i.e., more widely accessible) version of its flagship model. Called Abliterated Large Flash, the new model supports image and text inputs and comes with a context window of one million tokens, which is on par with some industry-leading models like OpenAI\u2019s GPT-5.6 Sol<\/a>. In a Friday X post, Abliteration described it as \u201ca cheaper alternative to our [previous] model<\/a> for product owners and defenders who need access to the most cyber-capable models.\u201d<\/p>\n Now live: abliterated-model-large-flash (our version of DeepSeek-V4.1-Flash-1017). It is a cheaper alternative to our GLM-5.3-based model for product owners and defenders who need access to the most cyber-capable models.<\/p>\n \u2013 1M context \u2014 Abliteration (@abliteration_ai) October 23, 2026<\/a><\/p>\n Cyber \u201cdefenders,\u201d however, might not be the only ones interested in putting the new model to use, especially now that the core architecture upon which much of the modern internet is based has been shown to be vulnerable to attack.<\/p>\n Albliteration\u2019s name is a reference to a technique used by AI researchers to alter or remove the mechanisms within an AI model\u2019s weights which would ordinarily prevent it from responding to certain kinds of requests. Those mechanisms have become a touchy subject in recent months, as many AI users have complained that frontier models are a little too sensitive with their safeguards<\/a>, often mistaking harmless requests for dangerous ones. Abliteration, in contrast, has positioned itself essentially as the anti-safeguard AI company: a more permissive alternative people can turn to if they\u2019re tired of AI telling them what they can and can\u2019t request. The larger model upon which Abliterated Large Flash is based \u201cdoes the offensive cyber, red teaming, and agent testing work other models refuse to do,\u201d the company said in another X post<\/a> when it was released back in August. (A TechCrunch report<\/a> published last month found Abliteration\u2019s AI granted requests to write Python code to steal Chrome passwords and cultivate a dangerous pathogen.)<\/p>\n While Abliteration clearly sees this permissiveness as something to celebrate, others in the industry have tried to sound the alarm on the removal of guardrails. <\/p>\n Late last month, Anthropic said in a report that its researchers were able to use abliteration and other techniques to breeze past the guardrails<\/a> built into GLM-5.3, a popular model from Chinese developer Z.ai (which also happens to power Abliteration\u2019s new model). The Anthropic researchers then ran a series of tests to see if the modified version of the model would refuse ethically dubious requests. It failed each one miserably. Just some relatively minor tweaks to the model\u2019s underlying weights, and almost all ethical boundaries suddenly seemed to dissolve. \u201cGiven this evidence,\u201d Anthropic wrote in its report, \u201cwe think it\u2019s likely both state and non-state actors will use models like GLM-5.3 to cause real-world harm.\u201d<\/p>\n Of course, even frontier, heavily safeguarded AI systems have their foibles. Anthropic itself has reported<\/a> that its own models have escaped sandboxes and hacked into third-party organizations, prompting the company to briefly pause<\/a> its internal tests. On Friday\u2014the same day Abliteration announced the launch of its cheaper new model\u2014Anthropic published<\/a> the results of an internal investigation showing that Claude Mythos 5, when tasked with a simple cybersecurity challenge earlier this year, found a code repository from GitHub belonging to actual human developers, tried to compromise it with malicious code, and \u201ccreated fake user identities that contacted the maintainer of the repository in an attempt to pressure them into approving the code change,\u201d according to Anthropic\u2019s report.\u00a0<\/p>\n Frontier AI models, as recent events have shown, don\u2019t need assistance from humans in order to commit cybercrime. But that calls for more safeguards and constraints around accessibility, not less.<\/p>\n","protected":false},"excerpt":{"rendered":"Cybersecurity experts are calling it \u201cBlack October.\u201d\u00a0 Earlier this week, researchers disclosed two major vulnerabilities hidden in PHP,…\n","protected":false},"author":2,"featured_media":639456,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[20],"tags":[365,6166,363,364,1510,18724,111,139,69,145],"class_list":["post-639455","post","type-post","status-publish","format-standard","has-post-thumbnail","category-artificial-intelligence","tag-ai","tag-anthropic","tag-artificial-intelligence","tag-artificialintelligence","tag-cybersecurity","tag-hacking","tag-new-zealand","tag-newzealand","tag-nz","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts\/639455","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/comments?post=639455"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/posts\/639455\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/media\/639456"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/media?parent=639455"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/categories?post=639455"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/nz\/wp-json\/wp\/v2\/tags?post=639455"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
\u2013 Vision
\u2013 Zero input\/output retention or\u2026<\/p>\n