Researchers Develop Tool to Protect Musicians from Deepfake Songs

My Music My Choice disrupts AI voice cloning to prevent unauthorized use of artists' vocals

Published on Mar. 5, 2026

Got story updates? Submit your updates here. ›

Researchers at Binghamton University, in collaboration with startup Cauth AI, have developed a tool called My Music My Choice that can protect musicians' songs from being cloned by AI-powered deepfake technology. The tool adds imperceptible changes to a song's waveform that prevent AI models from accurately replicating the vocals, while still allowing human listeners to enjoy the music as intended.

Why it matters

The rise of deepfake technology has enabled the creation of high-quality, studio-level imitations of artists' vocals, leading to issues around intellectual property rights, lost revenue, and emotional distress for musicians. My Music My Choice aims to empower artists by giving them a way to safeguard their work from this growing problem.

The details

My Music My Choice works by introducing small, undetectable modifications to a song's waveform. These changes confuse AI voice-cloning models, causing them to produce distorted noise instead of an accurate replica. However, the changes are imperceptible to human listeners, who will hear the original song as intended. The researchers have tested the tool on 150 music tracks across multiple genres and plan to continue expanding the testing to larger data samples.

  • The paper on My Music My Choice was presented at the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: AI for Music.

The players

Umur Aybars Ciftci

A research assistant professor in the First-Year Research Immersion Program at Binghamton University, who collaborated on the development of My Music My Choice.

Ilke Demir

The CEO and founder of Cauth AI, a startup company that collaborated with Binghamton University on the development of My Music My Choice.

Gerald Pena Vargas, Alicia Unterreiner, and David Ponce

Binghamton University students who contributed to the research on My Music My Choice.

Bad Bunny

A popular musician used as an example of an artist whose songs could be targeted by deepfake technology.

Got photos? Submit your photos here. ›

What they’re saying

“Even though this AI technology has been developed for fun and entertainment, a lot of people are using it for nefarious purposes. You can easily take someone's voice and make them sing something that they normally don't sing, or steal someone's songs and make it look like it is your song to begin with.”

— Umur Aybars Ciftci, Research Assistant Professor, Binghamton University (Mirage News)

“Collaborating with disruptive startups like Cauth AI provides us with a unique vantage point into the frontline challenges of the industry, essentially bridging the gap between lab-scale concepts and industrial-scale impact. Our goal is to build a model that figures out exactly which tiny modifications to introduce so that people hear no difference at all, while AI voice-cloning systems are thrown off.”

— Umur Aybars Ciftci, Research Assistant Professor, Binghamton University (Mirage News)

What’s next

The researchers will continue testing My Music My Choice on larger data samples and compare it to similar methods, though they note there are not many comparable tools available.

The takeaway

My Music My Choice represents a promising solution to the growing problem of deepfake songs, empowering musicians to protect their work from unauthorized AI-generated imitations while still allowing fans to enjoy the music as the artist intended.