Job Information

岗位背景:

我们正在寻找一位安全策略运营,负责公司AIGC产品内容安全策略优化与落地,覆盖多模态图文视频生成与智能Agent交互场景;你将与算法、业务团队协作,针对Prompt注入、深度伪造、隐形违规等新型AI风险,建设全链路防控能力。

岗位职责:

· 推进全链路安全策略建设与落地:针对Prompt注入、深度伪造、隐形违规等新型AI风险,设计并落地覆盖多模态图文视频生成与智能Agent交互场景的防控策略,平衡合规风控与用户体验;

· 优化智能审核规则与安全Prompt工程:持续迭代审核规则,独立评估策略效果,拆解漏拦、误拦Badcase,研判模型能力边界,联动算法团队提升风险识别精准度与场景覆盖率;

· 跟进风控数据监控与策略迭代:常态化关注拦截率、召回率、体验损耗等核心指标,完成案例复盘和结构化复盘报告,输出优化方案并推动策略迭代,实现风控能力闭环;

· 开展大模型红队测试与攻防沉淀:常态化开展红队测试与红蓝对抗,主动挖掘模型越狱、提示词攻击、隐性风险等漏洞,沉淀攻防案例库与防御策略,前置防控未知风险;

· 推进合规策略落地与突发事件处置:跟进AI行业合规监管政策,动态更新安全规则与应急处置流程,协同业务、算法团队推进落地,高效处置线上突发安全事件。

任职要求:

· 本科及以上学历,3年及以上AIGC内容安全、大模型风控、AI安全策略相关经验;

· 熟悉大模型基础原理与常见安全风险,掌握Prompt工程、模型对齐、红队测试等实操能力,具备风控策略设计、调优及落地经验;

· 熟悉国内生成式AI、网络内容安全相关法规标准,具备合规落地、风险治理及突发事件处置实战经验;

· 风险敏感度高,擅长案例拆解与归因分析,能够独立承接专项攻坚,跨团队推动问题闭环。

加分项:

· 有多模态产品安全治理经验;

· 有智能Agent安全治理经验。

About Meitu
Founded in 2008, Meitu is an AI-driven technology company that carries "beauty" as its core ideal. To "let art and technology converge elegantly", Meitu is developing outstanding products that streamline the production of image, video, and design, while striving to advance industry digitalization through related solutions.
Meitu leverages AI in the pursuit of its mission. In 2010, Meitu founded its R&D hub MT Lab, which is dedicated to AI and imaging innovations such as Computer Vision, Deep Learning, and Computer Graphics.
As of June 30th 2025, Meitu has 280 million monthly active users globally and 98 million monthly active users outside of Mainland China.
About the department
Our Corporate Functions team spans PMO, Finance, Legal, HR, and more. We foster an agile, supportive workplace through effective collaboration and system optimization, providing solid backing for business teams and steering the company forward.