研究表明,99.9%的人无法察觉深层假象,引起人们对身份欺诈的担忧。 Study reveals 99.9% of people can't detect deepfakes, raising concerns over identity fraud.

iProov的一项研究表明,99.9%的人无法准确地探测到深层假象,这些假象是人工智能生成的、真实的图像或视频。 A study by iProov shows that 99.9% of people cannot accurately detect deepfakes, which are AI-generated images or videos that look real. 研究强调,深假视频尤其难以发现,与图像相比,参与者正确识别假视频的可能性低36%。 The research highlights that deepfake videos are especially hard to spot, with participants being 36% less likely to correctly identify fake videos compared to images. 虽然 Tinder 和 Hinge 等约会网站上只有一小部分内容是经过 AI纵的，但深度伪造引发了人们对身份欺诈和错误信息的担忧。 While only a small fraction of content on dating sites like Tinder and Hinge is AI-manipulated, deepfakes are raising concerns about identity fraud and misinformation. 研究表明,需要改善生物鉴别安全和AI识字。 The study suggests a need for better biometric security and AI literacy.