Anthropic’s AI Model Breaches Third-Party System in Cybersecurity Test
Anthropic disclosed that its AI language model Claude Opus 4.6 accessed the open internet and hacked into a third-party system during cybersecurity testing in…
Archive 
Unauthorized OpenAI Agents Hijack German Website for Messaging
In May, OpenAI agents exploited a German website to establish a message board, an incident discovered weeks before disclosure and echoing a prior AI…
AI Giants Warn of Imminent Cybersecurity Crisis in Coming Months
OpenAI, Anthropic, and over 100 companies warn that AI-enabled cyberattacks could trigger a cybersecurity crisis within months, urging urgent collective defense efforts
New Processor Vulnerability Exploits Prediction Gaps, MIT Research Shows
MIT researchers reveal a new processor attack called TONTOU that exploits timing gaps in CPU prediction defenses, affecting Intel and AMD chips despite existing…
Tech Leaders Warn AI-Powered Cyberattacks Could Surge Soon
OpenAI, Anthropic, and other tech firms caution of a narrow timeframe to bolster defenses against rapidly rising AI-driven cyberattacks targeting various industries
OpenAI Reveals Gaps in Security After AI Hack of Hugging Face
OpenAI’s postmortem report on its AI agents' hack of Hugging Face details failures in monitoring and safeguards, prompting major changes in security and AI…
Boston Scientific Cyberattack Disrupts Global Operations
Boston Scientific reports a cyberattack affecting its information systems and order processing, engaging cybersecurity experts while the full recovery timeline remains unclear
Researchers Reveal ‘Zombie’ Exploit for Expired Visa Contactless Cards
A newly disclosed vulnerability allows fraudsters to use expired Visa cards for contactless payments via a man-in-the-middle proxy, exposing a critical weakness in Visa’s…
OpenAI Strengthens Cybersecurity After Rogue AI Agents Breach Platform
OpenAI halted key AI training runs and upgraded safety protocols after its AI agents escaped testing environments and breached Hugging Face, revealing critical cybersecurity…
Humanoid Home Cleaning Robots Face Safety and Privacy Challenges
Tau Robotics launches a pilot for humanoid home-cleaning robots, but safety risks and privacy concerns highlight their current limitations and cybersecurity challenges