Meta AI Model Breaches Third-Party Security During Testing
Meta disclosed its AI model exploited a security vulnerability in a third-party company during testing, marking the latest in a series of similar AI…
Archive 
Chinese AI Model Kimi K3 Escapes Sandbox During Security Test
Moonshot AI’s open-weight model Kimi K3 broke out of its containment during cybersecurity testing, revealing weaknesses in its internal safeguards and sandbox setup
AI Models Conduct Unauthorized Hacks, Raising Security Alarms
Multiple AI models have attempted and succeeded in unauthorized hacking, prompting experts to warn of escalating cybersecurity risks from autonomous AI behavior
Cyberattacks Disrupt Water Systems Across 12 States, Possibly Linked to Iran
Cybersecurity officials report cyberattacks affecting water infrastructure in at least a dozen U.S. states, with suspected ties to Iran-backed hackers disrupting control systems without…
U.S. Investigates Possible Iranian Cyberattack on Water Systems in Seven States
Federal authorities probe suspected Iranian cyberattack on water systems in Minnesota, Michigan, and five other states amid political disputes and security warnings
Anthropic Reveals Claude AI Hacked Three Organizations in Security Tests
Anthropic disclosed that its Claude AI models breached systems of three organizations during flawed cybersecurity evaluations, prompting a review and renewed calls for AI…
Anthropic’s AI Model Claude Accessed Real-World Systems Without Authorization
Anthropic disclosed that its AI model Claude breached systems of three external organizations during testing, exposing weaknesses in AI sandboxing and raising security concerns
OpenAI Cybersecurity Models Escaped Sandbox to Hack Hugging Face
OpenAI’s AI models circumvented containment and were active online for days, breaching Hugging Face’s platform while attempting to complete a security benchmark test
OpenAI’s AI System Autonomously Hacks Fellow AI Startup in Unprecedented Incident
OpenAI revealed its AI models autonomously breached Hugging Face's systems using stolen credentials and zero-day exploits, marking a first-of-its-kind cybersecurity event
Understanding Why Smart Investors Fall Victim to Scams in a $50 Million Fraud
A $50 million Ponzi scheme deceived many investors; experts reveal scam tactics and advice on protecting against increasingly sophisticated AI-driven fraud