Think Your Anonymous Account Is Safe? AI Might Figure Out Who You Are
大型语言模型可通过拼凑分散的公开线索,以惊人准确度解构伪匿名社交媒体账号。该技术利用碎片化信息识别用户真实身份,挑战了匿名账户的安全性假设。相关研究由美国斯坦福大学等机构开展,旨在揭示 AI 在去匿名化方面的能力边界。
EVENT DOSSIER
Recent research shows that large language models have the ability to de-anonymize pseudonymous social media accounts with remarkable accuracy by piecing together scattered public clues. This technology uses fragmented information to identify a user’s true identity, challenging the assumption of the security of anonymous accounts. Related research was conducted by institutions such as Stanford University in the United States, aiming to reveal the limits of AI’s capabilities in de-anonymization.
大型语言模型可通过拼凑分散的公开线索,以惊人准确度解构伪匿名社交媒体账号。该技术利用碎片化信息识别用户真实身份,挑战了匿名账户的安全性假设。相关研究由美国斯坦福大学等机构开展,旨在揭示 AI 在去匿名化方面的能力边界。