Meta公司近日宣布，自5月份起，其旗下的社交媒体平台Instagram、Threads以及Facebook将开始对“疑似由AI生成的内容”添加特殊的“水印标记”。这一举措旨在应对日益严重的AI产生的隐私和版权问题，科技行业的关注点正逐渐聚焦于此。
据Meta发布的新闻稿，公司将运用算法与人工审核的双重机制来识别“可能由AI生成的内容”。一旦内容被标记，用户在查看时将看到一个明显的水印，提示该内容可能并非由人类原创。同时，用户也有权主动在自己的AI生成图片上添加注释，声明其AI来源，这样的信息也将以水印形式出现在图片上，以便其他用户辨识。
Meta的这一决策反映了其对平台内容真实性和用户知情权的重视，也是对当前AI技术快速发展带来的挑战的一种应对。通过实施水印标记制度，Meta旨在维护其社交网络的透明度，同时提醒用户注意可能存在的虚假信息，从而加强社区的信任和安全。这一政策的实施效果及对用户行为的影响，无疑将受到业界和用户的密切关注。
英语如下：
News Title: “Meta to Implement New Policy in May: AI-Generated Content to be Marked with Watermarks”
Keywords: Meta, AI Watermark, Social Media
News Content: Meta recently announced that starting from May, its social media platforms, Instagram, Threads, and Facebook, will begin adding special “watermark labels” to content “suspected to be AI-generated.” This move aims to address the growing concerns over privacy and copyright issues related to AI-generated content, which is attracting increasing attention in the tech industry.
According to Meta’s press release, the company will employ a dual mechanism of algorithms and human review to identify content “likely created by AI.” Once marked, users will see a prominent watermark when viewing the content, indicating that it may not be human-originated. Additionally, users will have the option to proactively annotate their AI-generated images, declaring their AI source. This information will also appear as a watermark on the image, facilitating recognition by other users.
Meta’s decision reflects its commitment to the authenticity of platform content and users’ right to know. It constitutes a response to the challenges posed by the rapid advancement of AI technology. By implementing the watermarking system, the company aims to maintain transparency within its social network and alert users to potential misinformation, thereby enhancing community trust and safety. The impact of this policy on user behavior and its effectiveness will undoubtedly be closely watched by both industry insiders and users.
【来源】https://www.ithome.com/0/760/207.htm
Views: 2