Sony Develops AI Tool That Can Detect Copyrighted Songs in AI Music

The debate around AI-generated music and copyright just took a significant turn. Sony Group has developed new technology that can identify original songs used to train artificial intelligence systems and detect their influence in AI-created tracks.

As concerns grow over how AI models are trained, especially in the music industry, this new development could reshape conversations around AI music copyright, artist compensation, and transparency in machine learning.

Here’s what you need to know.

According to reports from Nikkei Asia, Sony’s research division, Sony AI, has built a system capable of analysing AI-generated songs and identifying which original tracks were used during training. More importantly, it can estimate how much influence each source song had on the final output.

In simple terms, if an AI tool creates a song, Sony’s technology can examine it and trace its roots. It can detect whether elements of copyrighted music were used to train the system and determine the level of similarity or influence.

This is a major step in the growing debate around AI training data transparency and the use of copyrighted material in machine learning models.

The rise of AI music generators has been rapid. Tools powered by large language models and generative AI can now compose songs that mimic existing artists, genres, and production styles with striking accuracy.

However, many of these systems are trained using massive datasets that may include copyrighted songs. 

This is not the first time generative AI has forced a reckoning around originality and ownership. The same tensions have already surfaced in visual art, where creators have challenged AI companies over how their works were used in training datasets as explored in my earlier article titled: With AI Remixing Everything, Can Anything Still Be Called Original?. The central question has remained consistent: if AI systems learn from copyrighted material at scale, who owns the output?

That unresolved question now extends to music. If an AI system learns from thousands of existing songs, who should be paid when it produces a track influenced by them? 

Artists, record labels, and rights holders have raised serious concerns about:

• Unauthorised use of copyrighted music

• Lack of compensation for creators

• Difficulty proving infringement

• Unclear training data sources

What makes Sony’s development significant is that it moves the debate beyond theory. Instead of arguing abstractly about influence, the company claims it can technically identify which original songs contributed to an AI-generated track, and to what extent. In a landscape where lawsuits have often centred on opaque training data, the ability to trace influence could shift discussions from accusation to measurable evidence.

Sony’s new AI music detection tool reportedly works in two main ways.

If AI developers agree to collaborate, Sony can directly access information about which songs were used in the training process. This makes attribution straightforward.

If developers do not cooperate, the system can analyse an AI-generated track and compare it against Sony’s music catalogue. Using advanced pattern recognition, it estimates which original songs contributed to the final result and how strongly they influenced it.

This approach shifts the focus from guessing to measurable analysis. Instead of arguing about similarities, the technology aims to quantify influence.

For the music industry, that level of detail could prove valuable.

For musicians and rights holders, the core concern has always been compensation.

If an AI model trains on copyrighted material and later produces songs that resemble those works, artists want recognition and payment. Without a way to track influence, that becomes difficult.

Sony’s technology opens the door to:

• Revenue-sharing frameworks

• Improved copyright enforcement

• Greater transparency in AI model training

• Clearer licensing agreements

Rather than blocking AI-generated music outright, this system could allow for a structured compensation model. Artists whose songs influenced AI output might receive payment based on measurable contribution.

That approach may be more practical than banning AI tools altogether.

Beyond music, this development signals something larger. As generative AI expands into writing, art, film, and design, similar concerns arise around training data and copyright. Systems capable of tracing influence may become essential across industries.

The ability to measure how much one work contributes to another could reshape discussions around intellectual property in the AI era.

For now, Sony’s innovation focuses on music. However, its broader implications are clear: transparency and accountability are becoming central to AI development.

RELATED: PlayStation’s February 2026 State of Play: All the Big Game News You Need to Know

Sony’s new AI music detection technology represents a significant development in the ongoing conversation about AI-generated music and copyright protection.

By identifying original songs within AI outputs and estimating their influence, the system could support fairer compensation models and strengthen intellectual property enforcement.

As artificial intelligence continues to reshape the creative landscape, solutions that balance innovation with creator rights will become increasingly important.

The music industry has entered a new phase, one where artificial intelligence and copyright law are no longer separate conversations, but deeply connected ones.