#red teaming

37 stories · 5 in 7 days · 12 in 30 days
Stories, products and papers tagged "red teaming", newest first.
Nov 13
Chinese contentMobileProbe:基于轨迹感知注入的移动智能体安全评估框架

做手机 GUI 智能体或相关安全工作的话,这篇值得看:它不靠人工写死攻击载荷,而是顺着智能体实际执行轨迹自动找注入点,83 个实例覆盖 12 个应用,攻击成功率比人工和零样本 LLM 基线都高,可以直接当部署前审计工具用。

arXiv cs.AI@Yingbin Jin et al. (10)Source
Nov 12
Chinese contentMicrosoft Research 播客解析红队、蓝队与紫队网络安全实践

Microsoft Research 的人聊 AI 时代怎么做安全测试,把红队、蓝队、紫队各自干什么讲得很清楚。做安全或想给自家系统做攻防演练的工程师可以听听,比单纯讲攻防对抗的视角更完整。

Event
Official accountMicrosoft Research@MSFTResearchSource
PapersBuildersSelected02:34
Perplexity Red-Teams SPACE VM Sandbox with 12 Models

The non-obvious insight is that the CTF format itself may bias agent security evaluations, meaning current red-team benchmarks could be testing the wrong thing. Teams building or deploying computer-use agents, sandbox infrastructure, or agentic safety pipelines should read the methodology to audit their own evaluation setups. It also suggests enterprises cannot rely on a model's security-specialization claims as evidence of sandbox resistance.

Official accountPerplexity@perplexity_aiSource
Nov 7
Nov 1
Oct 27
Oct 24
Oct 15
ProductsBuildersMulti-sourceSelected12:51
White Circle open-sources RISE red-teaming tool for image models

The shift from red-teaming as an internal lab practice to a reproducible open-source workflow lowers the barrier for any team to stress-test image model safeguards themselves. Builders and safety engineers can plug in their own models and run the same attacks that broke frontier safeguards. It also illustrates a broader trend of safety tooling being engineered with commodity models rather than bespoke training runs.

Event
elvis@omarsar0Source
ProductsBuildersSelected05:48
White Circle releases RISE for testing image generator safety

The credit-assignment design matters: by rewarding strategies across many generated prompts, RISE keeps finding attacks even against heavily filtered models, a common failure point for red-teaming tools. Safety teams running image generation services can deploy it directly against their own models since both code and attacker models are public. It contrasts with closed audit tools by making the full attack search reproducible.

rohanpaul_ai@rohanpaul_aiSource
Oct 12
Oct 9
Oct 8
Oct 1
ProductsBuildersSelected00:45
LangSmith Engine v2 adds red teaming

This proactive testing approach could significantly reduce production failures in AI applications by identifying vulnerabilities earlier in development. Developers building complex AI agents can now implement continuous red teaming within their workflow, potentially saving hours of debugging time and improving application reliability.

Official accountLangChain@LangChainAISource
Sep 28
Sep 11
Sep 6
Aug 19
Aug 13
Jul 29
Jul 17
Jul 16
Jul 14
Jul 9
Jul 3
Jun 23
Jun 9
Jun 7
Jun 4
May 29
PapersSelected72°13:00
Chinese content400次实验揭示LLM攻击一致性:Gemini 2.5 Flash-Lite成功率85%领先

这项研究揭示了LLM作为攻击者的行为规律和可靠性差异,做AI安全评估或红队测试的团队值得关注——它告诉你不同模型在真实攻击场景下的稳定性和失败模式,直接指导模型选型和防御策略。

arXiv: Anthropic@Galip Tolga ErdemSource
May 12