[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-anthropic-blocks-cyber-bio-and-chemistry-queries-on-fable-5-model":10},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":24,"persona_id":22,"persona_name":22,"section":22,"tags":25,"sources":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},505,"anthropic-blocks-cyber-bio-and-chemistry-queries-on-fable-5-model","Anthropic blocks cyber, bio and chemistry queries on Fable 5 model","On June 9, Anthropic added safety filters to Fable 5, refusing any prompts about cybersecurity, biological weapons or chemical synthesis.","Anthropic has tightened Fable 5’s safety guardrails.\n\nOn June 9 the company announced new restrictions for its latest frontier model. Fable 5 will now refuse any request that touches three areas: cybersecurity tactics, biological‑weapon research and chemical‑synthesis instructions. Anthropic said the topics are “too dangerous” for open conversation and that the filters are active by default. The change was rolled out to all API users without a separate opt‑in.\n\nThe move sharpens Anthropic’s public safety posture and puts pressure on competitors to justify similar limits. Developers building security‑aware tools will need workarounds or a different model, which could slow some use cases. Regulators have praised the step as aligning with emerging AI risk guidelines.\n\nIt’s a reminder that “open” AI models still come with built‑in roadblocks, even if the marketing brochures suggest otherwise.","[\"anthropic\",\"ai-safety\",\"large-language-models\"]","2026-06-09T19:20:25.000Z","2026-06-09T23:21:57.266Z","2026-06-09T23:22:55.075Z","published",null,[],"https:\u002F\u002Fcdn.xyz.onl\u002Farticle-images\u002Fanthropic-blocks-cyber-bio-and-chemistry-queries-on-fable-5-model.webp",[26,27,28],"anthropic","ai-safety","large-language-models",[30],{"name":31,"url":32},"Ars Technica","https:\u002F\u002Farstechnica.com\u002Fai\u002F2026\u002F06\u002Fanthropic-says-these-topics-are-too-dangerous-to-let-its-fable-5-model-talk-about\u002F",0]