The AI firm Deeptrace found 15,000 deepfake videos online in September 2019, a near doubling over nine months. Facebook Twitter Pinterest A comparison of an original and deepfake video of Facebook chief executive Mark Zuckerberg. But deepfakes themselves were born in 2017 when a Reddit user of the same name posted doctored porn clips on the site. Poor-quality deepfakes are easier to spot. But they also pose a personal security risk: deepfakes can mimic biometric data, and can potentially trick systems that rely on face, voice, vein or gait recognition.

Source: The Guardian January 13, 2020 09:56 UTC