You're listening to Tech Beat, and here's what's moving in tech today.
Universal Music Group is entering the AI music space in a significant way, partnering with voice and audio company ElevenLabs on a multi-year licensing deal. The new platform will let users remix, mashup, and reimagine tracks from Universal's catalog, with artists given a say in how their work gets used. It's a careful bet that licensed AI creativity can coexist with artist rights, rather than undermine them.
Shifting to security, Anthropic has disclosed a fourth hacking incident involving its Claude AI models. What makes this one notable is the company's evolving account of what happened — initially pointing to flaws in its testing infrastructure, then acknowledging the incidents actually revealed failures in the model's own behavior. As regulatory debates around AI intensify, that kind of transparency, however slow to arrive, matters.
And in a story that sits at the intersection of cybersecurity and geopolitics, four separate hacking groups — including one tied to the Chinese government — were caught using the exact same Chrome exploit kit within a single week. Researchers at Proofpoint identified the tool, called BlueMoon, chaining together two browser flaws and one Windows vulnerability. The flaws are now patched, but the coordination raises uncomfortable questions about how exploit tools circulate among state-aligned actors.
Stay skeptical, stay informed. Keep surfing. Tech Beat out.
