CERT/CC reveals systemic jailbreak flaws in major AI models, including ChatGPT and Claude, enabling bypass of safety filters with identical prompt techniques.
AI security
NVIDIA discloses 3 vulnerabilities in the NeMo Framework enabling remote code execution and data tampering. Update to version 25.02 to secure your AI development.
CVE-2024-0132: This report details how an incomplete NVIDIA Container Toolkit patch allows container escape and denial-of-service (DoS) attacks, affecting AI systems and Docker on Linux
Invariant Labs reveals a critical vulnerability in the Model Context Protocol (MCP) enabling Tool Poisoning Attacks. Learn how this threat can lead to data exfiltration and AI hijacking
Google introduces Unified Security, powered by Gemini AI, to streamline enterprise cybersecurity. Discover new AI security agents, enhanced threat detection, and data protection tools for comprehensive security management.