National Trading Standards has published its first watchlist detailing how criminals use AI to deceive shoppers, including fake retailer websites styled as UK family businesses showing Union Jack-themed High Streets. Victims who order from these sites often receive nothing, or get cheap mass-produced goods from countries like China instead of the advertised UK-made products. The watchlist also flags voice cloning and deepfake videos as growing tools used in fraud.
bbc.co.uk
· 2026-10-18
A Tokyo court found that a TikTok account used AI to replicate the voice of veteran anime actor Kenjiro Tsuda in narration videos posted between July 2024 and late 2025, ruling this violated his publicity rights. The court declined to order removal of the videos since the anonymous account had already been deleted. Tsuda's legal team had argued the account's 200,000-plus followers were drawn by hearing a recognizable star's voice narrate lurid content, while TikTok countered the voice was merely generic.
engadget.com
· 2026-10-01
Fideuram chairman Paolo Molesini was deceived by a fake WhatsApp message posing as Intesa Sanpaolo CEO Carlo Messina and a follow-up call using an AI-generated voice impersonating a law firm partner. Believing the request was legitimate, Fideuram wired €95 million to accounts in China and Hong Kong before staff spotted irregularities. Authorities in Italy, Portugal and China helped recover €53 million, and Molesini has since resigned.
techspot.com
· 2026-09-28
ElevenLabs released two new speech models, v4 and v4 Turbo, adding expanded expression control, faster voice cloning from just 10 seconds of audio, and support for over 90 languages, up from 70. The company said quality improved most notably in Japanese, Brazilian Portuguese, Mandarin and Cantonese, and the new models offer lower latency aimed at voice agent applications.
techcrunch.com
· 2026-09-28
Google has released Gemini 3.8 Flash TTS and Flash-Lite TTS, new text-to-speech models that can generate custom voices from text descriptions, replicate a voice from a 30-second sample with permission, and follow directions for accent, pacing, emotion and multi-speaker dialogue. The models support over 100 languages and more than 2,000 ready-made voices, and are rolling out in Gemini Notebook and Google Vids as well as to developers via Google AI Studio and the Gemini API.
androidauthority.com
· 2026-09-24
New research indicates that scammers using AI-generated voices that mimic a target's own accent are more effective at deceiving victims than generic synthetic voices. The findings suggest that accent-matching adds a layer of perceived familiarity and trust that makes fraudulent calls more convincing.
wsj.com
· 2026-09-13
More than 80 performers, including Matt Lucas, Hugh Bonneville, Siobhán McSweeney and singer Sandi Thom, have signed a letter urging the UK government to introduce legislation protecting performers from unauthorised AI voice cloning. The signatories argue current law fails to safeguard an artist's voice, face and identity as AI technology advances. The letter follows earlier warnings from legal experts that existing UK rules are outdated and cannot fully protect personal likeness or voice.
bbc.co.uk
· 2026-08-28
GoKawiil