Acclaimed composer Max Richter has argued that human composers should be paid when artificial intelligence systems use their work to make music, saying that today's AI systems were built on data that was taken without permission.
Max Richter · AI, music and fair pay
Composers should be paid when AI learns from their music
Richter argues that today's AI music systems were built on material taken without permission, and that the human composers whose work made training possible should be compensated when those systems generate new tracks.
How the argument flows
Step 1
Existing music
Composers' works are the raw material.
→
Step 2
AI training
Models learn from that material, which Richter says was not licensed.
→
Step 3
New tracks
Generated music is produced at scale.
→
Richter's ask
Payment to composers
The people whose work enabled training are paid.
The case for payment
Training relied on existing compositions.
That material was not licensed or paid for.
Creators whose work made the models possible should share in the returns.
Clearer rules are needed on how copyrighted works are used.
The counterargument
Some AI developers say training on publicly available material is lawful.
They also argue that training is transformative rather than copying.
The broader debate over consent and royalties remains unsettled.
Key statement
Principle is clear. Practice is not yet defined.
The reporting does not yet answer the questions that will decide whether Richter's position becomes practice:
How a compensation scheme would work
Who would pay
How individual contributions would be measured
Whether any AI company has agreed to such terms
Legal status of training on copyrighted music, by jurisdiction
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article...