AI-Generated Music: Challenges in Copyright Protection

Kiran Sharma


Modern AI is more sophisticated and has a far greater impact, as direct creation tools have replaced indirect improvements. For instance, Google’s Magenta programme created a new “Nirvana” song by studying the lyrics, guitar riffs, chord progressions, and melodies of the group’s earlier songs. Text instructions are given to ChatGPT, which then creates lyrics that are better than those IBM Watson created for Alex da Kid in 2016. AI-powered artists are available for leasing through Authentic Artists. A 300-question personality test is used by to generate personalized compositions. Song starter from Bandlab is a “AI-powered idea generator” that can produce royalty-free music in a matter of seconds. Staccato, a start-up, bills itself as “an AI Lennon to your McCartney” because it can bounce song writing concepts off human writers. So many illustrations of AI-generated music will lead to many questions also specially with respect to copyright legal framework of these work. This article focuses on the ownership of AI-generated work, the question of originality in such works and if protected, who will be held responsible in case of any infringement?

Key Words:Artificial Intelligence (AI), Music, Copyright

Recommended Citation

Kiran Sharma, ‘AI-Generated Music: Challenges in Copyright Protection’ (2023) 1(1) OUJIPR, 161 accessed [date] Available at:

download PDF