London-based startup Surf Security has launched a beta version of a groundbreaking browser equipped with AI deepfake detection. This tool claims up to 98% accuracy in identifying whether online voices are human or AI-generated, using advanced neural network technology.
Surf’s system, powered by State Space Models, analyzes audio inconsistencies across languages and accents to detect cloned voices in under two seconds. It supports a range of platforms, including WhatsApp, Slack, Zoom, and Google Meet, and features built-in background noise reduction for better accuracy.
Deepfake audio poses an increasing threat, as evidenced by recent fake clips of Sir David Attenborough. These technologies have been exploited for fraud, misinformation, and reputational damage. Surf’s browser aims to protect enterprises, media, and law enforcement from these risks.
Surf also plans to incorporate AI image detection into its toolkit and expects to release the full version next year.
News Source:thenextweb，This article does not represent our position.
