About our Deepfakes news
Latest news on deepfakes, covering deepfake videos, deepfake audio, voice cloning, face swap apps, AI-generated images, synthetic media, deepfake scams, deepfake detection, non-consensual deepfake images, election disinformation, generative AI and deepfake laws.
Deepfakes are images, videos or audio created or altered with artificial intelligence to make someone appear to say or do something they did not. The name blends deep learning and fake. Common techniques include face swaps, lip-syncing and voice cloning, and the results can be hard to tell apart from genuine recordings.
Deepfakes are now used in fraud, politics and abuse. In March 2022 Meta removed a fake video of Ukraine's President Zelensky urging his troops to surrender, which had appeared on a hacked news site. In February 2024 the US Federal Communications Commission (FCC) ruled that AI-generated voices in robocalls are illegal, after a fake call imitating President Biden reached voters before the New Hampshire primary. In 2024 the engineering firm Arup confirmed that an employee in Hong Kong had been tricked on a video call with digitally recreated colleagues, in a fraud worth about 25 million US dollars.
Detection tools analyse faces, voices and file data for signs of manipulation, but their accuracy varies and studies show that people often struggle to spot fakes. Proposals for content provenance include digitally signing media at the point of capture. Practical advice includes checking the source, looking for visual glitches and confirming unusual money requests through a second channel.
Victims include public figures, women and girls targeted with fake intimate images, and companies targeted by fraud. UK law has criminalised sharing sexually explicit deepfakes under the Online Safety Act and, more recently, creating them, while the US passed the TAKE IT DOWN Act in 2025. Campaigners argue that platforms and AI developers should do more to prevent misuse, while others warn that broad rules could restrict satire, art and free expression.
Techniques for faking images and film are old, but the modern deepfake emerged in late 2017, when a Reddit user called deepfakes posted face-swapped videos made with machine learning. Generative adversarial networks, which emerged in the mid-2010s, were a key turning point. Reddit banned the community in early 2018 under its policy on involuntary sexual content.
Seeing is no longer always believing, so context matters more than ever. Our NewsNow deepfakes news feed brings together the latest headlines on deepfakes, voice cloning, synthetic media, scams and detection from reliable sources worldwide, so you have a comprehensive, up-to-date view.
