The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has issued a binding directive that mandates the use of an 'AI Bill of Materials' (AIBOM) for all AI systems used by federal agencies a
AIBW Security DeskSynthesis AI, a startup specializing in AI-powered synthetic data generation, confirmed it was the victim of a sophisticated cyberattack. Threat actors breached the company's cloud infrastructure and
AIBW Security DeskA new research paper from Stanford University's AI Lab details a novel jailbreak technique named 'Cognitive Dissonance'. The attack forces a large language model to internally hold two contradictory,
AIBW Security DeskIn an effort to standardize security for AI applications, Google has released 'Model Guard', an open-source framework designed to protect against common LLM-based threats. Model Guard acts as a progra
AIBW Security DeskHealthLeap AI, a prominent provider of AI-driven medical diagnostic tools, disclosed a critical data breach that has exposed the sensitive records of approximately 2 million patients. The breach occur
AIBW Security DeskA new research paper from Stanford University's AI Lab has detailed a sophisticated adversarial attack named 'Contextual Weaving'. This technique bypasses the safety alignment of major large language
AIBW Security DeskA critical security breach has hit the popular AI library Litellm. Two versions on PyPI contained malicious code designed to steal sensitive API keys and environment variables.
AIBW Security DeskA stunning report from Le Monde shows France's aircraft carrier was tracked in real time using a consumer fitness app. This highlights a critical OPSEC failure, demonstrating how AI-powered data analysis can turn innocuous data into potent military intelligence.
AIBW Security DeskOpenAI reveals its security framework for designing AI agents resistant to prompt injection. This multi-layered approach constrains risky actions and requires user consent.
AIBW Security DeskOpenAI's latest research introduces the Instruction Hierarchy Challenge, a new training method to make LLMs more resistant to prompt injection and safer to deploy.
AIBW Security DeskOpenAI is acquiring AI security platform Promptfoo, signaling a major push to embed robust vulnerability detection and remediation directly into the AI development lifecycle.
AIBW Security DeskOpenAI has launched Codex Security, a new AI agent in research preview. It promises to revolutionize application security by detecting, validating, and patching vulnerabilities.
AIBW Security Desk