［1］包泽芃, 钱铁云. 大模型红队测试研究综述［J］. 计算机科学, 2025, 52(1): 3441［2］李南, 丁益东, 江浩宇, 等. 面向大语言模型的越狱攻击综述［J］. 计算机研究与发展, 2024, 61(5): 11561181［3］Shir T, Sagi T. Wiz Research finds architecture risks that may compromise AIasaService providers and consequently risk customer data; works with Hugging Face on mitigations［EBOL］. 2024 ［20250916］. https:www.wiz.ioblogwizandhuggingfaceaddressriskstoaiinfrastructure［4］新智元. 第一个被人类骗钱的AI傻了, 近5万美元不翼而飞！ Scaling Law还能带我们到AGI吗?［EBOL］. (20241130) ［20241213］. https:mp.weixin.qq.comsfKA4cO1VvvnWqSsTdsM_MA［5］秦臻, 庄添铭, 朱国淞, 等. 面向人工智能模型的安全攻击和防御策略综述［J］. 计算机研究与发展, 2024, 61(10): 26272648［6］OWASP Group. OWASP Top 10 for large language model applications［EBOL］. 2023 ［20240913］. https:owasp.orgwwwprojecttop10forlargelanguagemodelapplic