AI-Infra-Guard
PulseAugur coverage of AI-Infra-Guard — every cluster mentioning AI-Infra-Guard across labs, papers, and developer communities, ranked by signal.
-
GitHub AI infrastructure projects surge in popularity
This week's trending projects on GitHub focus on the underlying infrastructure for AI agents rather than the agents themselves. Tools like Maka, llmfit, AI memory, AI-Infra-Guard, and vLLM Semantic Router are gaining tr…
-
Tencent study finds DeepSeek Harness vulnerable to prompt injection; Correctover offers solution
A security assessment by Tencent's AI-Infra-Guard team revealed that DeepSeek Harness is vulnerable to indirect prompt injection attacks, with success rates ranging from 17% to 25.5% across various channels. The study c…
-
Tencent paper reveals 25.5% indirect prompt injection success in DSH
A recent security paper from Tencent's AI-Infra-Guard team highlights significant vulnerabilities in the DeepSeek Harness (DSH) open-source agent orchestration framework. The research found that indirect prompt injectio…
-
DeepSeek Harness vulnerable to indirect prompt injection, study finds
Researchers have evaluated the security of the DeepSeek Harness framework against indirect prompt injection attacks. The study, conducted using the AI-Infra-Guard (A.I.G) system, involved over 14,500 controlled executio…
-
New framework audits vendor-hosted LLM APIs for quality degradation
Researchers have developed Ventor-QTest, a novel black-box auditing framework designed to verify the quality of inference APIs for vendor-hosted large language models. This method employs both repeated-request and long-…
-
Three open-source AI security tools compared: Garak, AI-Infra-Guard, Vigolium
This article compares three open-source AI security tools: garak, AI-Infra-Guard, and vigolium. Garak, developed by NVIDIA, focuses on scanning LLM responses for vulnerabilities like jailbreaks and data leakage. AI-Infr…