日内瓦会议后，AI治理必须直面信任缺口，深入探讨全球协作与公众信心重建。
Article URL: https://www.techpolicy.press/after-geneva-ai-governance-must-confront-the-trust-deficit/ Comments URL: https://news.ycombinator.com/item?…
MCP服务器安全评分工具上线，CI集成与采购审查一步到位。
As a security professional whose employer has given Claude and various other AI tooling to everyone at the company, I kept finding myself as the bottl…
最新调查显示，大多数护士认为AI在患者护理上还不够可靠，医疗AI信任问题再引关注。
Article URL: https://www.washingtonpost.com/technology/2026/07/07/most-nurses-say-ai-isnt-good-enough-trust-with-patient-care-survey/ Comments URL: ht…
追踪开源聊天LLM十二个版本的信任度变化，揭示模型行为随时间漂移的潜在风险
arXiv:2607.02587v1 Announce Type: cross Abstract: Model cards quote trust-benchmark scores without recording when they were measured, and the same num…
扩散与流模型的黑盒对齐新方法，信任区域噪声搜索带来高效微调
arXiv:2603.14504v2 Announce Type: replace-cross Abstract: Optimizing the noise samples of diffusion and flow models is an increasingly popular approac…
将信任区域优化引入策略蒸馏，解决模型压缩中的策略漂移问题。
arXiv:2607.04751v1 Announce Type: cross Abstract: Big goals are hard to achieve all at once; breaking them into small steps is wiser. We present Trust…
警惕AI大公司利用你的研究IP打压竞争，别把核心研发数据交给它们。
I am very paranoid about sharing potential AI research with e.g. Claude [Code] or ChatGPT/Codex. I believe that any company is essentially a paperclip…
探讨AI agent在持续部署中的信任控制局限，提出渐进授权思路，打破“全有或全无”困境。
AI coding agents are showing up in infrastructure workflows. They can diagnose a failed terraform apply , summarise what changed across a dozen module…
AI编程助手Cursor让开发者放手，无人工审查也能直接上线生产环境，代码生成信任度飙升
IT之家 6 月 29 日消息，据 Business Insider 报道，AI 编码智能体正越来越多地获得信任，可在无人工监督的情况下独立完成工作。 人工智能编程助手 Cursor 的最新数据显示，过去半年里，无需额外人工审核、直接上线生产环境的 AI 生成代码变更占比大幅飙升。 这一趋势表明，开…
用贝叶斯锚定潜在信任区域优化高维不确定下的结构设计分类问题，突破传统贝叶斯优化局限
arXiv:2604.25241v2 Announce Type: replace Abstract: Categorical structural optimization under aleatoric uncertainty is challenging because each design…
新方法Trust Region Masking，通过动态约束策略更新提升大模型长序列强化学习稳定性与收敛性。
arXiv:2512.23075v5 Announce Type: replace-cross Abstract: Policy gradient methods for Large Language Models optimize a policy $\pi_\theta$ via a surro…
这篇论文提出“城堡与守门人”传递信任理论，为第三方网络安全风险治理提供了全新分析框架
arXiv:2606.26866v1 Announce Type: cross Abstract: Third-party vendors, such as analytics platforms, cloud services, identity providers, and software s…
一条销售公式揭露了信任破碎时再多价值也白搭，值得每个做业务的人细品。
This is an equation i invented to help me communicate how sales work to the team...it's simple but trust that nobody can hack it; even the unicorns: o…
零信任本地优先的RAG ETL引擎，用Rust+WASM+Apache Arrow在浏览器沙箱里完成全部管道，安全又高效。
Most RAG pipelines ship raw, sensitive documents over the wire to cloud services just to get them parsed, scrubbed of PII, chunked, and vectorized. Bi…
编程代理的“虚假成功”比失败更危险——即使通过测试，也可能完全偏离意图，威胁生产环境信任度。
arXiv:2603.25764v3 Announce Type: replace-cross Abstract: As coding agents move into production workflows, teams need to know not only whether an agen…
论文揭示不可靠反馈如何破坏LLM代理的工具使用能力，提醒不要盲目信任反馈机制。
arXiv:2606.21409v1 Announce Type: new Abstract: Tool-augmented agents are typically evaluated by their gains under reliable external feedback. Yet the…
想在时代广场打广告但嫌贵？来拼单分摊，10秒只需100美元，低成本建立品牌信任。
Hey, Planning to put an ad on NYC Times-square billboard to build trust. Buying a 60sec ad and splitting into 6 different company ad slots turns out t…
从非标量信任出发，探讨在代理链中携带类型化退化信息并传播，提升链式系统的可靠性与透明度。
Two posts ago, in the one about agents failing quietly , I handed you a fix for silent degradation: tag a degraded output trust="degraded" , propagate…
探索商业场景中人类与AI代理的交互模式，实证揭示协作效率与信任建立的底层逻辑。
arXiv:2606.18716v1 Announce Type: cross Abstract: As AI agents are increasingly integrated into core business processes, understanding and designing e…
Arch用户因信任风险从AUR转向官方仓库，引发包管理安全性的社区讨论
I'm spending a lot of my time removing AUR packages for alternatives in the official Archlinux repositories. I've shifted from Dropbox to RClone, from…