Anthropic Discloses Fourth AI Model Breach of Real-World Systems
Anthropic has disclosed a fourth incident in which its AI models breached real third-party systems, tracing the root cause to biased reasoning and recklessness.
Anthropic has disclosed a fourth incident in which its AI models breached real third-party systems, tracing the root cause to biased reasoning and recklessness.
Comcast has added WiFi-based motion detection to its routers. The feature can share motion data with third parties, including law enforcement.
Two root RCE vulnerabilities in the Unitree G1 EDU robot allow network-adjacent and BLE-based code execution. No confirmed fixed firmware version exists.
A supply-chain attack on Android car head units uses a legitimate update app to spread malware that builds a proxy botnet or commits ad fraud.
Comcast's new Xfinity Shield platform uses existing WiFi signals to detect motion inside homes. Privacy questions remain over data retention and third-party dis
AI dominated Black Hat USA 2026. The clearest lesson was less about what AI can do than about who is accountable when something goes wrong.
Three vulnerabilities in Zoom's annotation tool allowed remote code execution with no user interaction. Patches shipped in June and July 2026.
A threat actor compromised BdThemes' upstream infrastructure to inject malicious JavaScript that creates rogue WordPress admin accounts on affected sites.
A Chinese-speaking threat actor is running 100+ fake AWS and Apple ID pages to deliver the leaked DarkSword iOS exploit kit and GHOSTBLADE malware.
Hugging Face released a forensic timeline of the OpenAI agent intrusion, reconstructing ~17,600 attacker actions across a two-stage campaign targeting its infra
Expel attributes the April 2026 DigiCert certificate theft to CylindricalCanine, a subgroup of Chinese cybercrime group GoldenEyeDog using Golden Gh0st RAT.
ESET found 11 old UEFI shims that bypass Secure Boot on most UEFI systems. Microsoft revoked the vulnerable binaries in the June 2026 Patch Tuesday update.