Deepfakes & AI-Generated Media

In late 2017 a Reddit user posting under the handle "deepfakes" began sharing face-swapped pornography of celebrities, grafting famous faces onto the bodies of adult performers using a free machine-learning technique. The handle gave the whole emerging field of AI-generated synthetic media its name. Alarm grew as doctored clips of world leaders and ordinary people fuelled fears that anyone could be convincingly impersonated. Europol, the EU's law-enforcement agency, published a 2022 report titled Facing Reality? warning of an "information apocalypse" in which society could "no longer distinguish fact from fiction." Activist Kate Isaacs described the trauma of discovering her face weaponised in fake explicit videos.

The panic accelerated in 2024. AI-generated explicit images of Taylor Swift went viral on X, reportedly reaching 45 million views before the platform suspended search for her name. In Hong Kong, scammers used a deepfake video call of a chief financial officer to trick staff at engineering firm Arup into transferring $25 million. In South Korea, a Telegram-based deepfake sexual-abuse epidemic exposed victims at over 500 schools. An AI-generated robocall impersonating President Biden hit New Hampshire voters before the 2024 Democratic primary, urging them not to vote.

Regulation followed. The UK Online Safety Act 2023 criminalised sharing deepfake intimate imagery; a 2025 amendment added criminal liability for creating it. The US NO FAKES Act, which would give individuals a federal right to sue over unauthorised AI replicas of their voice or likeness, was reintroduced in 2024–25. And EU AI Act Article 50 — which obliges providers and users of generative AI to clearly label synthetic audio, images and video so audiences know what they are seeing — becomes enforceable from August 2026. What began as a niche Reddit curiosity has become a mainstream regulatory concern in under a decade. What is being protected — the integrity of public speech and the dignity of a person's likeness — is older than the technology attacking it; the rules are being rebuilt in real time, and the outcome is still open.