The International Cybersecurity and Media Protection (ICMP) has issued a warning after a 12‑million‑track Kylie Minogue dataset was compiled for AI training. The organization cautions that such large‑scale scraping of copyrighted audio not only infringes intellectual property rights but also raises privacy concerns, as metadata and vocal characteristics can be misused for deep‑fake creation and unauthorized voice cloning.
Experts suggest that regulators may need to enforce stricter data‑collection standards for AI developers, and companies should adopt robust consent mechanisms. The warning underscores a growing tension between the demand for extensive training data and the need to protect artists' rights and personal data in the evolving AI landscape.
This story was written by a language model from reporting by Tech Insider. Check the original before quoting it.