用检索增强大模型读取直播间的“似曾相识”线索，跨会话证据让风险预判更聪明
arXiv:2601.16027v2 Announce Type: replace Abstract: The rise of live streaming has transformed online interaction, enabling massive real-time engageme…
关注Claude AI模型带来的金钱安全隐患，揭示AI如何放大网络攻击能力。
Article URL: https://www.rte.ie/brainstorm/2026/0517/1572343-claude-ai-model-security-issues-humans-work-information-data/ Comments URL: https://news.…
新研究揭示AI模型在武装冲突地区部署时，可能因对齐失败而无意中激化矛盾，对AI安全与全球治理敲响警钟。
arXiv:2605.22720v1 Announce Type: new Abstract: AI models are already deployed in societies affected by armed conflict, and journalists, humanitarian …
8500万美元基金瞄准60万亿美元受灾房地产，气候风险催生私人资本新机遇
After launching to invest in fire tech, Convective is broadening its mandate to disaster resilience.
新方法保障LLM在线部署每轮风险可控，基于共形预测与RLVR训练，安全认证更可靠。
arXiv:2605.20270v1 Announce Type: new Abstract: A local specialist LLM, fine-tuned with reinforcement learning from verifiable rewards (RLVR) on opera…
AI图谱技术首次系统揭示肥胖对全身器官的隐藏损伤，颠覆传统认知。
Article URL: https://medicalxpress.com/news/2026-05-ai-atlas-reveals-hidden-body.html Comments URL: https://news.ycombinator.com/item?id=48212931 Poin…
K8s调度LLM虽顺手，但隔离不足的默认配置暗藏工具调用与数据泄露风险，云原生社区必须重新审视信任边界。
TL;DR: Kubernetes schedules LLM workloads well, but it does not give them the isolation boundary they need once they start calling tools, executing co…
基于真实开发者查询，揭示LLM生成代码中库幻觉的具体风险与数据支撑
arXiv:2509.22202v3 Announce Type: replace-cross Abstract: Large language models (LLMs) now play a central role in code generation, yet they continue t…
系统评估AI艺术生成中风格泄露现象，揭示生成模型模仿艺术风格时的版权与伦理风险。
arXiv:2605.17500v1 Announce Type: new Abstract: Generative text-to-image models are typically trained on large-scale web-scraped datasets that include…
揭示记忆增强的LLM代理在长期使用中面临的安全风险，提醒关注AI系统持续记忆带来的新威胁。
arXiv:2605.17830v1 Announce Type: cross Abstract: Safety evaluations of memory-equipped LLM agents typically measure within-task safety: whether an ag…
探讨大语言模型能否作为信用风险模型的事后可解释工具，前沿研究视角值得关注。
arXiv:2602.18895v2 Announce Type: replace-cross Abstract: Large language models (LLMs) have shown promise in translating model-based explanations into…
多智能体市场中LLM代理集体行为如何引发经济风险？这篇论文提出系统性对齐方案。
arXiv:2605.17698v1 Announce Type: new Abstract: The deployment of Large Language Models (LLMs) as autonomous economic agents introduces systemic risks…
揭秘SaaS产品与保险的结构相似性，一个商业建模新视角。
arXiv:2605.16699v1 Announce Type: new Abstract: Capped-usage SaaS products -- LLM subscriptions such as Claude Code and ChatGPT, cloud platforms such …
OpenAI联合多机构发布AI恶意使用预测报告，提出预防与缓解威胁的路径。
We’ve co-authored a paper that forecasts how malicious actors could misuse AI technology, and potential ways we can prevent and mitigate these threats…
大模型对齐新方法：通过风险约束解码感知人类偏好分歧，提升模型鲁棒性。
arXiv:2603.08145v2 Announce Type: replace Abstract: Preference-based alignment methods (e.g., RLHF, DPO) typically optimize a single scalar objective,…
探讨非平稳环境下加权风险最小化的快速收敛性，理论分析与新界证明，适合机器学习理论研究者。
arXiv:2602.05742v2 Announce Type: replace-cross Abstract: Weighted empirical risk minimization is a common approach to prediction under distribution d…
前OpenAI员工警告xAI安全记录不佳，或成SpaceX IPO的潜在障碍
The ex-employees, who cofounded a new AI watchdog group, say investors deserve more information about xAI’s safety practices before SpaceX goes public…
新框架DriveSafe专为驾驶场景风险检测与安全建议设计，提升行车智能安全。
arXiv:2605.16892v1 Announce Type: cross Abstract: Comprehensive situational awareness is essential for autonomous vehicles operating in safety-critica…
Hacker News 社区热议：该全职创业还是继续打工？社区成员分享经验与权衡。
I want to ask HN because it feels this community is more tech oriented then most out there. I have been building an idea for the past couple of months…
同时使用多个AI编码助手（Codex、Copilot、Claude Code）是否存风险？Hacker News上的真实讨论。
Hello People, I am using codex and copilot , claude code at the same time for my projects.Does it have potential risks for using all of them ? Comment…