As AI-generated music continues to flood streaming platforms and social media, researchers at University of Chicago say they may have found a way to help listeners tell what’s real and what’s artificial, according to CBS Chicago.
A team from the university’s SAND Lab has developed a browser extension called “Quicksilver,” designed to identify whether a song was likely created using artificial intelligence. The tool reportedly works by scanning audio for hidden patterns and digital artifacts commonly left behind by AI music generators such as Suno and Udio.
Researchers say AI-generated songs are becoming increasingly difficult for everyday listeners — and even music professionals — to identify without technical assistance. According to the team, some streaming platforms are now seeing massive increases in AI-created uploads, raising concerns about transparency, copyright, and the future of human artists in the music industry.
Unlike some detection systems that require uploading audio files to outside servers, Quicksilver reportedly runs directly on a user’s device, allowing songs to be analyzed in real time while protecting privacy. Developers say the goal is not to stop AI music completely, but to provide more transparency for listeners and creators as the technology rapidly evolves.
The project was led by prominent UChicago researchers Ben Zhao and Heather Zheng alongside graduate student Stanley Wu. The team says the extension could eventually help streaming services, record labels, and listeners better understand how much AI-generated content is circulating online.
The release of the tool comes as debates surrounding AI-generated art, music, and entertainment continue to grow worldwide. Many artists have voiced concerns about AI systems being trained on copyrighted material without permission, while others argue the technology could reshape the future of creativity entirely.
For now, Quicksilver represents one of the first major attempts to give listeners a way to spot AI-made music before pressing play.
