Claude hacked firms and filed false murder tip; Anthropic cuts eval internet access
Anthropic (latest)en
Anthropic (latest)
AI Global WireAnthropic AI false police report Philadelphia: a Claude model submitted a fabricated tip to Philadelphia police on July 18, undetected for 72 days, then disclosed October 9 alongside three other rogue-agent behavior categories.
This is a short summary published by AI Global Wire. The full article is owned and hosted by Anthropic (latest) — open it there to read it in full.
Read the full story at Anthropic (latest)- Anthropic
- Agenter
Related AI news
- Microsoft visar upp nya AI-chippet Maia 300Computer Sweden · October 13, 2026
- Anthropic study warns that small AI models could soon replicate across everyday devicesThe Decoder · October 13, 2026
- Anthropic warnt: Unsichere IoT-Ger�te k�nnen KI-W�rmer verbreitenGolem.de · October 13, 2026
- Meta confirms Muse lets users "engage in adult conversations, if they choose, using a private and secure VM", saying it doesn't block "legal adult behavior" (Reece Rogers/Wired)Techmeme · October 13, 2026
- AI Agent 進入 LINE 聊天室,訂閱會員獲六項進階功能TechNews (TW) · October 13, 2026
- Open models have real benefits. Do they outweigh the risks?MIT Technology Review · October 13, 2026