OpenAI揭露了一场与中国国家有联系的AI驱动运动,目的是骚扰在国外的持不同政见者,包括假告状和有协调的虚假信息,从而导致用户被禁止使用并引起全球对AI滥用的关注。 OpenAI exposed a Chinese state-linked AI-driven campaign to harass dissidents abroad, including fake obituaries and coordinated disinformation, leading to user bans and global concern over AI misuse.

OpenAI发现中国使用AI工具(如ChatGPT), 计划并记录恐吓海外持不同政见者的努力, 包括伪造假告、散布虚假信息、协调假社交媒体账户等。 OpenAI has uncovered a global Chinese state-linked operation using AI tools like ChatGPT to plan and document efforts to intimidate dissidents abroad, including fabricating fake obituaries, spreading disinformation, and coordinating fake social media accounts. 调查的焦点是一位中国执法人员, 他利用AI作为数字日记追踪针对批评者, 包括日本首相高一桑娜(Sanae Takaichi)的竞选活动, The investigation centered on a Chinese law enforcement official who used the AI as a digital journal to track campaigns targeting critics, including Japan’s Prime Minister Sanae Takaichi, though ChatGPT refused some requests. OpenAI核查了真实世界的影响, 如假死亡报告和协调在线攻击, 并禁止用户使用。 OpenAI verified real-world impacts, such as false death reports and coordinated online attacks, and banned the user. 调查结果揭示了涉及数百名特工和先进数字战术的大规模、系统的努力,突出表明人们日益关切在跨国镇压和地缘政治影响运动中滥用人工智能的问题。 The findings reveal a large-scale, systematic effort involving hundreds of operatives and advanced digital tactics, highlighting growing concerns over AI misuse in transnational repression and geopolitical influence campaigns.