1
Anthropic 联合创始人奥拉：人工智能的发展不能仅由科技公司主导
AI发展需要多方共治，Anthropic联合创始人呼吁监管不应只交给科技公司。
IT之家 5 月 25 日消息，Anthropic 联合创始人克里斯・奥拉（Chris Olah）于当地时间周一表示，人工智能的发展不能完全交由科技公司，他呼吁宗教领袖、政府和民间社会加强监督。 IT之家注意到，奥拉在教皇利奥十四世（Pope Leo XIV）首份关于人工智能的通谕发布仪式上发言称，…
2
MAGIQ: A Post-Quantum Multi-Agentic AI Governance System with Provable Security
后量子时代多智能体AI治理系统，提出可证明安全的MAGIQ架构，解决新兴计算范式下的安全挑战。
arXiv:2605.06933v2 Announce Type: replace Abstract: Our computing ecosystem is being transformed by two emerging paradigms: the increased deployment o…
3
亿格云完成数亿元B轮融资，加码“人+AI”统一安全治理
亿格云获数亿元B轮融资，主攻“人+AI”统一安全治理新赛道，连续三年翻倍增长。
4
When AI override paths become production paths
AI事故的真正根源往往不是模型本身，而是架构中未受审核的“覆盖路径”悄然成为生产路径，值得每个技术管理者反思。
Article URL: https://heavythoughtcloud.com/blog/when-the-override-path-becomes-the-production-path Comments URL: https://news.ycombinator.com/item?id=…
5
Formal Methods Meet LLMs: Auditing, Monitoring, and Intervention for Compliance of Advanced AI Systems
形式化方法与LLM结合，为AI系统合规提供审计、监控和干预的全生命周期治理方案。
arXiv:2605.16198v1 Announce Type: new Abstract: We examine one particular dimension of AI governance: how to monitor and audit AI-enabled products and…
6
Our approach to AI safety
OpenAI官方阐述AI安全策略，从构建到部署全流程的安全管控框架。
Ensuring that AI systems are built, deployed, and used safely is critical to our mission.
7
Practices for Governing Agentic AI Systems
OpenAI发布代理AI系统治理实践，为构建安全可控的自主AI提供关键框架。
8
Built to benefit everyone
OpenAI重组强化使命治理，扩大资源以负责任推进AI发展，造福全人类。
OpenAI’s recapitalization strengthens mission-focused governance, expanding resources to ensure AI benefits everyone while advancing innovation respon…
9
Governance of superintelligence
OpenAI前瞻思考超级智能治理，为未来AI系统监管奠定基础
Now is a good time to start thinking about the governance of superintelligence—future AI systems dramatically more capable than even AGI.
10
AI Knows When It's Being Watched: Functional Strategic Action and Contextual Register Modulation in Large Language Models
AI在被观察时会刻意调整语言风格？这项研究用百场多智能体辩论实验证实了LLM存在“霍桑效应”：不仅区分人类与AI观察者，甚至对AI审计系统也有适应性反应，对AI治理与审计提出全新挑战。
arXiv:2605.15034v1 Announce Type: cross Abstract: Large language models (LLMs) have been extensively studied from computational and cognitive perspect…