https://aim-intelligence.inblog.io Security Layer for Trustworthy AI Agents Mon, 25 Nov 2024 03:52:58 GMT inblog Mon, 19 May 2025 22:52:50 GMT https://image.inblog.dev?url=https%3A%2F%2Finblog.ai%2Fapi%2Fog-logo%3FlogoUrl%3Dhttps%253A%252F%252Finblog.ai%252Finblog_logo.png%26title%3DAIM%2520Intelligence%26description%3DSecurity%2520Layer%2520for%2520Trustworthy%2520AI%2520Agents&w=1920&q=85 https://aim-intelligence.inblog.io https://aim-intelligence.inblog.io/quicktake-oneshot-is-enough-consolidating-multiturn-attacks-into-efficient-singleturn-prompts-for-llms-55771 null Mon, 19 May 2025 22:54:39 GMT https://aim-intelligence.inblog.io/quicktake-oneshot-is-enough-consolidating-multiturn-attacks-into-efficient-singleturn-prompts-for-llms-55771 team team https://aim-intelligence.inblog.io/elite-enhanced-languageimage-toxicity-evaluation-for-safety-quicktake-55700 ELITE introduces a rubric-driven way to judge how Vision–Language Models (VLMs) handle malicious multimodal prompts, then packages those judgements into a large, well-balanced benchmark. The team shows that popular “refusal-rate only” metrics over-estimate jailbreak success, while their toxicity-aware rubric tracks human annotations far better. Mon, 19 May 2025 03:06:00 GMT https://aim-intelligence.inblog.io/elite-enhanced-languageimage-toxicity-evaluation-for-safety-quicktake-55700 team team https://aim-intelligence.inblog.io/exploiting-mcp-security-threats-llms Discover how attackers exploit vulnerabilities in the Model Context Protocol (MCP) to manipulate Large Language Models (LLMs), steal data, and disrupt operations. Learn real-world attack scenarios and defense strategies to secure your AI systems. Fri, 09 May 2025 00:00:00 GMT https://aim-intelligence.inblog.io/exploiting-mcp-security-threats-llms team team https://aim-intelligence.inblog.io/36482 For the Pursuit of Safe and Trustworthy AI Wed, 27 Nov 2024 06:44:55 GMT https://aim-intelligence.inblog.io/36482 team team