OpenAI发布治理蓝图，以强化网络安全与前沿AI防御。2026年10月21日，OpenAI发布一份政策蓝图，为政府和行业提出建议，旨在让网络安全防御者应对新兴人工智能威胁。
OpenAI Publishes Governance Blueprint to Strengthen Cybersecurity and Frontier AI Defenses On October 21, 2026, OpenAI published a policy blueprint outlining proposals for governments and industry to empower cybersecurity defenders against emerging artificial intelligence threats.
OpenAI认为，在先进进攻性AI工具广泛传播之前，存在一个狭窄的“防御者窗口期”，并提出三项支柱措施：加强前沿模型的保障与关闭控制、现代化网络防御和关键基础设施，以及通过独立技术评估和强制性事件报告扩大公共机构监督。
Arguing that a narrow 'defender's window' exists before advanced offensive AI tools spread widely, OpenAI proposed measures across three pillars: strengthening safeguards and shutdown controls for frontier models, modernizing cyber defenses and essential infrastructure, and expanding public institutional oversight through independent technical evaluations and mandatory incident reporting.