{"rewrite":{"id":"r_6817ba3816394fb555418da8","clusterId":"c_dd4a2e5df9b4f8ae169c9cea","slug":"anthropic-study-finds-ai-cyberattacks-are-already-automated","model":"deepseek-v4-flash","headline":"Anthropic Study Finds AI Cyberattacks Are Already Automated","summary":"Anthropic analyzed 832 suspended accounts over a year and found that AI use in cyberattacks is shifting from simple preparation to complex post-compromise activities. The proportion of attackers classified as medium risk or higher rose from 33% to 56% over the study period. Traditional risk indicators no longer correlate with attacker skill.","whyItMatters":"The study shows AI is lowering the technical barrier for advanced cyberattacks, making traditional threat assessment methods obsolete.","webCardHtml":"\u003cp\u003eAnthropic examined 832 accounts suspended for cyberattack activity between March 2025 and March 2026, mapping them to the MITRE ATT\u0026CK framework. The most common AI use was preparation, with 67.3% of accounts using AI for tasks like malware creation. But the study found a clear shift: AI-assisted phishing dropped 8.6% while AI use for identifying valid accounts inside compromised networks rose 8.9%. Attackers classified as medium risk or higher jumped from 33% in the first six months to 56% in the second six months.\u003c/p\u003e\u003cp\u003eAnthropic also reported that traditional risk indicators no longer work. The least skilled attackers used an average of 16 techniques, while the most skilled used about 20, and the platform used showed no correlation with risk level. The company concluded that AI is automating advanced attack stages that previously required technical expertise.\u003c/p\u003e","blueskyPost":"Anthropic's year-long study of 832 suspended accounts shows AI cyberattacks are already automated. Post-compromise techniques are rising, phishing is falling, and traditional risk scoring is failing.","twitterPost":"Anthropic's study of 832 suspended accounts over a year finds AI cyberattacks are automated. Post-compromise techniques are rising, phishing is falling, and old risk indicators no longer work.","threadsPost":null,"newsletterBlurb":"Anthropic analyzed 832 accounts suspended for cyberattacks over a year and found that AI use is shifting from simple preparation to complex post-compromise activities. The proportion of attackers classified as medium risk or higher jumped from 33% to 56%. Traditional risk indicators no longer correlate with attacker skill.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20260604-ai-enabled-cyber-threats/\",\"title\":\"Anthropic reports that 'cyberattacks are already automated with AI'\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":4170,"outputTokens":557,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1780546829,"createdAt":"2026-06-04T04:12:56.000Z","publishedAt":"2026-06-04T04:16:19.000Z","updatedAt":"2026-06-04T04:16:19.000Z"},"cluster":{"id":"c_dd4a2e5df9b4f8ae169c9cea","canonicalTitle":"Anthropicが「サイバー攻撃はすでにAIで自動化されている」と報告","representativeArticleId":"a_b519e7ab3283d604515b9b56","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":false,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[],\"studios\":[],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-06-04T02:58:00.000Z","lastSeenAt":"2026-06-05T06:10:00.000Z","updatedAt":"2026-06-05T06:18:40.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20260605-anthropic-ai-build-itself/","title":"Anthropicが「AIがAIを作る」自己改善ループのリスクを警告、AI開発をAI自身が加速する可能性を論じる"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":[],"studios":[],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":["Anthropic analyzed 832 accounts suspended for cyberattack activity between March 2025 and March 2026.","The proportion of attackers classified as medium risk or higher rose from 33% to 56% over the study period.","AI-assisted phishing dropped 8.6% while AI use for identifying valid accounts inside compromised networks rose 8.9%.","The least skilled attackers used an average of 16 techniques, while the most skilled used about 20.","Traditional risk indicators no longer correlate with attacker skill, according to the study."]}