Researchers from Stanford's AI Lab have published a paper detailing a novel jailbreak technique called Multi-Persona Prompt Injection (MPPI). The attack involves instructing an LLM to adopt multiple,
In a major move to standardize AI security practices, the US Cybersecurity and Infrastructure Security Agency (CISA) and the National Institute of Standards and Technology (NIST) have jointly launched
Researchers at Carnegie Mellon University have published a paper detailing a novel jailbreak technique named 'Semantic Blindspot.' This method effectively bypasses the safety and alignment filters of
SynthCore AI, a leading MLOps platform, has disclosed a critical security breach affecting thousands of its customers. The incident, which occurred in late June, was traced back to a sophisticated sup
The U.S. National Institute of Standards and Technology (NIST) has officially released the AI Secure Supply Chain Framework (AI-SSCF), a landmark policy designed to mitigate risks from data poisoning
AI safety and research company Anthropic confirmed it suffered a significant security breach. Unauthorized actors gained access to their internal cloud environment and exfiltrated proprietary data, in
Google's AI security division has announced the open-source release of 'Model Guardian,' a comprehensive framework for red teaming and auditing AI systems. This Python-based tool is designed to help d
Cloud services provider SynthAI confirmed a critical security breach today, resulting from a misconfigured cloud storage bucket. The incident exposed the proprietary model weights for its flagship 'Ne
The non-profit AI Sentry Foundation has launched Guardian v1.0, an open-source framework designed to help developers secure their LLM-powered applications. Guardian acts as a programmable firewall for
The Open Web Application Security Project (OWASP) has officially released GARNET (Generative Adversarial Red-teaming Network), a new open-source framework designed to automate and standardize the secu
Researchers from Stanford University's AI Lab have published a paper detailing a novel jailbreak technique named 'Contextual Weaving'. The attack bypasses safety alignments in large language models by
Cognition AI, the developer of the popular AI-powered code assistant 'CodeWeaver', announced a significant security breach. Attackers exploited a vulnerability in a third-party logging service integra