Lawyers square off in fight over voice data used to train AI

A wave of class-action lawsuits filed in federal court targets some of the biggest names in tech, alleging they vacuumed up voice recordings from journalists, podcasters, and voice actors to train AI systems without ever asking permission. The companies named read like a who’s who of Silicon Valley: Google, Amazon, Apple, Microsoft, Meta, Nvidia, Samsung, Adobe, and AI voice startup ElevenLabs. The suits, filed between May 12 and May 14 in the US District Court for the Northern District of Illinois, invoke Illinois’ Biometric Information Privacy Act, better known as BIPA. That law requires companies to get written consent before collecting biometric identifiers. The plaintiffs argue that voiceprints qualify as biometric data, no different from fingerprints or facial scans. The plaintiffs and the legal theory Among those bringing the claims is Carol Marin, a well-known Chicago journalist, alongside Pulitzer Prize winners Yohance Lacour and Alison Flowers. The legal theory is straightforward. BIPA was designed to ensure people control their own biometric data. Voiceprints, which the filings describe as immutable “digital fingerprints,” fall squarely under that umbrella. If a company ...