AI安全与可信评价专题研讨会 Seminar on AI Safety and Trustworthy Evaluation
面向人工智能安全、可信技术应用和评价方法建设，组织专家围绕风险识别、评审维度和责任技术实践开展专题研讨。 A focused seminar on AI safety, trustworthy technology applications, risk identification, evaluation dimensions, and responsible technology practice.
活动导语
人工智能安全与可信评价正在成为数字科学评价体系的重要方向。本次研讨会聚焦AI安全风险识别、可信应用评价和责任技术实践。
活动要点
- 梳理AI系统在数据、模型、应用和治理层面的主要风险。
- 讨论可信AI评价指标、材料要求和专家评审关注点。
- 交流安全贡献类成果的呈现方式与评价边界。
主要内容
会议将围绕模型安全、数据安全、可解释性、隐私保护和实际部署中的责任机制展开讨论，帮助参与者理解安全可信成果如何被专业、审慎地描述和评价。
参与对象
AI安全研究人员、数据治理人员、企业技术负责人、评审专家和关注可信技术应用的会员代表。
后续安排
研讨成果将作为后续AI安全相关活动、奖项申报说明和研究报告选题的重要参考。
Introduction
AI safety and trustworthy evaluation are becoming important parts of the digital science assessment framework. This seminar focuses on risk identification, trustworthy application evaluation, and responsible technology practice.
Key Points
- Map major AI risks across data, models, applications, and governance.
- Discuss trustworthy AI indicators, material requirements, and review concerns.
- Exchange approaches for presenting AI safety contribution outcomes.
Main Topics
The seminar will address model safety, data security, explainability, privacy protection, and accountability mechanisms in deployment, helping participants describe and evaluate trustworthy outcomes with professional care.
Participants
AI safety researchers, data governance professionals, enterprise technical leads, review experts, and member representatives focused on trustworthy technology application.
Follow-up
Insights from the seminar will inform future AI safety events, award-application briefings, and research report topics.