{"data":{"id":2356,"lang":"en","title":"Anthropic Says It Blocked State-Backed Attempts to Use Claude for Bioweapons Research","summary":"Anthropic's new threat report says it blocked five bioweapons-research attempts and a Russia-linked malware campaign that misused Claude between December 2025 and August 2026.","category":"ai","category_label":"AI","published_at":"2026-09-11T22:01:22Z","updated_at":"2026-09-11T22:01:22Z","url":"https://techtools.cz/tools/news/article/2356-anthropic-says-it-blocked-state-backed-attempts-to-use-claude-for-bioweapons","api_url":"https://techtools.cz/news-api/v1/articles/2356?lang=en","markdown_url":"https://techtools.cz/news-api/v1/articles/2356.md?lang=en","image_url":"https://coincentral.com/wp-content/uploads/2026/09/Anthropic-Claude-1-scaled.jpg","word_count":190,"reading_time_minutes":1,"sources":[{"url":"https://coincentral.com/anthropic-blocks-attempts-to-use-claude-ai-for-bioweapons-cyberattacks-and-propaganda","domain":"coincentral.com"},{"url":"https://www.tomshardware.com/tech-industry/artificial-intelligence/anthropic-says-claude-thwarted-bioweapon-research-from-state-sponsored-actors-covert-accounts-used-u-s-proxies-to-attempt-to-engineer-deadlier-viruses-tried-to-evade-identification-and-regional-blocks","domain":"tomshardware.com"}],"translation":{"lang":"cs","id":2429,"title":"Anthropic tvrdí, že zablokoval pokusy státem podporovaných aktérů zneužít Claude k výzkumu biologických zbraní","url":"https://techtools.cz/tools/news/clanek/2429-anthropic-tvrdi-ze-zablokoval-pokusy-statem-podporovanych-akteru-zneuzit-claude","api_url":"https://techtools.cz/news-api/v1/articles/2429?lang=cs"},"content":"Anthropic published a threat intelligence report on September 11 detailing misuse of its Claude models between December 2025 and August 2026, disrupted before causing harm. The company identified five separate attempts by researchers to use Claude for work that could support biological weapons development, including a request to help draft a grant application for gain-of-function research on the chikungunya virus aimed at increasing its transmissibility and immune evasion.\n\nThe report also documents a Russia-linked espionage group, Midnight Blizzard, allegedly using Claude to build malware that automatically rewrote itself whenever security tools detected it, plus attempts by the hacking group ShinyHunters and China-based labs. Anthropic separately uncovered nine coordinated influence operations originating from Russia, Iran, Turkey and other regions, using hundreds of fake social media accounts to push political messaging across six continents.\n\nAnthropic said it blocked every identified case, tightened safeguards on its newer Claude models, and shared intelligence with government authorities and industry partners. The findings add to a broader debate this week among AI safety researchers, some of whom have said the risk of advanced AI systems causing mass casualties within the next decade is a growing concern."}}