Indirect prompt injection is the most widespread and serious vulnerability in AI agents today, not just a theoretical risk. Research shows attacks can transfer across models and behaviors, revealing a fundamental weakness in how agents interpret context. More capable models aren’t safer, high performance often comes with equally high vulnerability. Attacks are especially dangerous because […]
Logitech reports attackers stole data likely belonging to customers, employees, and suppliers. The breach stemmed from a zero-day flaw in a third-party software platform that has since been patched. Logitech has not disclosed which data types or which platform were affected. The Cl0p ransomware group claims it exploited an Oracle E-Business Suite vulnerability to access […]