联合王国规定,在人工智能产生的虐待材料不断增加的情况下,必须及早进行人工智能检测,以检测儿童的安全情况。 UK to mandate early AI testing for child safety, amid rising AI-generated abuse material.

联合王国计划颁布一项法律,允许指定的儿童保护团体在早期开发时测试AI系统,以防止产生儿童性虐待材料,填补目前只有共享有害内容才能消除的空白。 The UK plans to introduce a law allowing designated child protection groups to test AI systems early in development to prevent creation of child sexual abuse material, addressing a gap where harmful content is currently only removed after being shared. 这项举措是"犯罪与警务法案"的一部分, 旨在从一开始就将安全性嵌入人工智能, The move, part of the Crime and Policing Bill, aims to embed safety into AI from the start, with oversight by experts to prevent data leaks. 大赦国际提供的关于儿童性虐待材料的报告在2025年增加了一倍多,达到426个案件,占所有非法在线内容的56%,而女孩作为目标的比例特别高。 Reports of AI-generated child sexual abuse material more than doubled in 2025, reaching 426 cases, making up 56% of all illegal online content, with girls disproportionately targeted. 政府官员和儿童安全倡导者欢迎这一步骤,但敦促所有开发商强制进行测试,以确保有效保护。 Government officials and child safety advocates welcome the step but urge making testing mandatory for all developers to ensure effective protection.