欧盟委员会周五在布鲁塞尔推出新的监管团队，以加强对全球AI公司的监督，这一举措恰逢新《AI法案》即将于周日（2025年1月26日）生效之际[1]。欧盟AI Office的团队规模扩编至38人[1]，将专门监测AI模型是否被用于发布色情内容、生成虚假图像和视频以及实施网络威胁等违规用途[1]。
作为执法措施的配套，欧盟建立了举报工具和合规工具供业界使用[1]。根据新规，AI公司需要使用标签或数字水印对生成内容进行标注[1]。对于违反规定的企业，欧盟有权对其处以罚款或禁止其进入欧盟市场[1]。欧盟技术主权首席Henna Virkkunen表示，这些新规代表了"朝着人们和企业能理解和信任的AI迈出重要一步"[1]。
The European Union unveiled a dedicated enforcement team in Brussels on Friday to monitor and combat the misuse of artificial intelligence systems, coinciding with the imminent implementation of the bloc's groundbreaking AI Act.[1] The newly expanded EU AI Office, bolstered by 38 additional staff members, will focus on detecting violations including the generation and distribution of non-consensual intimate imagery, synthetic deepfakes, and cyber threats.[1] The enforcement effort comes as the EU's comprehensive AI legislation is set to take effect on Sunday, January 26, 2025.[1]
Under the new regulatory framework, AI companies operating within the EU must label or apply digital watermarks to all generated content to ensure transparency and traceability.[1] The EU has established reporting mechanisms and compliance tools to identify violations, with enforcement powers that include imposing significant fines or restricting non-compliant companies' market access within the bloc.[1] Henna Virkkunen, the EU's chief technology sovereignty officer, characterized the enforcement initiative as "an important step toward AI that people and enterprises can understand and trust."[1]