Anthropic disclosed in July that a review of 141,006 cybersecurity evaluation runs had uncovered three incidents, spanning ...

Experts got Claude to port a PLC exploit, but it cost $536 and 8 hours, and a later AI-generated payload accidentally ...

Hackers are exploiting a critical Langflow flaw that lets unauthenticated attackers remotely execute Python code on ...

Threat actors are exploiting an unauthenticated remote code execution vulnerability (CVE-2026-0768) in Langflow, an ...

Fire Ant hackers, a China-linked espionage group, hacked Cisco routers and enterprise authentication servers in 2026, ...

Anthropic reward hacking research confirms flawed RL training produced Hacker-Opus, an AI model that attacked real systems ...

Tracked as CVE-2026-0768, the security defect allows unauthenticated attackers to execute arbitrary Python code remotely.

AI agents have escaped testing environments, communicated with one another and acted in unexpected ways. Experts say we should expect more incidents.

Anthropic admits Claude hacked three real companies in safety tests, then revealed a model trained to cheat, forge grades, ...

The big news today is of course the successful launch and deployment of NASA’s Nancy Grace Roman Space Telescope earlier this morning. The space agency’s latest observatory lifted off a 7:26 ...

Claude Code Auto Mode can run malware via a malicious ZIP despite safety checks. Read what the exploit reveals ...

OpenAI agents hacked Hugging Face in a 700-strong swarm after 1,200 agents exchanged 70,000 messages. ETIH’s edtech news report explains how.