“We dance through the night so bright, hand in hand feeling alive, our hearts ignite.”
These lyrics are sung in a raw, soulful voice which occasionally hits blue notes. The acoustic guitar provides rhythmic support, adding tasteful touches to complement the vocals. The song is titled ‘Rhythm of love’ and is both lively and heartfelt.
However, there's no human behind the voice, no hands on the electric guitar—there isn't even a guitar. The song has been created by Suno AI, a generative artificial intelligence music programme designed to produce lifelike songs.
The creator of the song is 54 years old Deepankar Sharma ‘Anwar’, a resident of Delhi, who saw in Suno AI an affordable conduit to preserve his poetry and retain ownership through a permanent medium of expression.
Anwar, who has created around 10 songs on the platform till now, came across Suno AI just weeks ago surfing through the Internet. He has been uploading these songs on his newly created YouTube channel ‘Bollywood AI Music’ which has 150 odd subscribers.
He told Decode, “Writing poems has always been my hobby. I was mulling over publishing them into a book but that is not necessarily very pocket-friendly. With Suno AI, I just need to pay $8 a month to subscribe to its annual premium version, which will transfer the copyright of the songs I create, to me.”
Currently Anwar is using the free version and hence the songs he creates are owned by Suno AI. “I want to shift to the paid version because I plan to monetise the creations. There is a chance that Suno AI can issue a copyright strike as I don’t own the songs,” he said.
Talking about the process of creating songs on Suno AI, Anwar explained that one can either feed in their own lyrics into the AI system, like him, or prompt the system to generate lyrics.
He further explained, “One just needs to understand the coding to get the tune right. As per their choice, a user can create a hip-hop song or a Bollywood 90s melody, by designing the desired initial instrumental tune, bridge between verses and hook of the song.”
Anwar is an alumnus of Delhi University's Zakir Hussain College, where his interests in poetry and the Urdu language were nurtured. He subsequently pursued an MBA from Madurai Kamaraj University and completed certification courses in programming and web design.
Currently working as a financial consultant, he previously traded in the stock market, but his involvement ended abruptly after incurring significant losses during Covid.
The 54-year-old feels that he has managed to create “decent” songs for his poems after multiple attempts on the platform. He said, “Thanks to AI, I believe this is the right time for me to pursue my passion for poetry and music.”
Though Anwar owes a lot to AI and technology, the self-admitting RD Burman fan does not feel that a melody can have a long shelf-life without the human touch.
“Creating music traditionally requires significant investments in instruments and studio time, but AI has significantly reduced those costs for me. Despite this, I believe that relying too heavily on technology can never produce the best results,” he said.
Suno AI- The New Buzz For Budding Musicians
Like Anwar, many others around the world have been using AI to either reduce costs or simplify their lives.
Founded by a Cambridge-based team of Michael Shulman, Georg Kucsko, Martin Camacho, and Keenan Freyberg, Suno AI has been made widely available since December 20, 2023. On March 21, 2024, the AI platform released its v3 version, allowing all users to create a limited number of 2-minute songs with a free account.
The company recently announced that over 12 million people have used its platform, globally, in the less than a year since its launch.
Over the past year alone, generative AI has made major strides in producing credible text, images and even video. However the companies providing these services have also been embroiled in lawsuits over copyright. The latest one to be added on the list is Suno AI.
Last month, the Recording Industry Association of America (RIAA) announced that labels, including Sony Music Entertainment, Universal Music Group Recordings, and Warner Records, have filed lawsuits against Suno AI and Udio for copyright infringement.
The lawsuits alleged that these AI music startups are exploiting the recorded works of artists ranging from Chuck Berry to Mariah Carey.
It was further claimed that Suno and Udio’s software stole music to “spit out” similar work. The lawsuit also asked for a compensation of $150,000 (Rs 1.25 crore) per work.
RIAA’s chairman and chief executive, Mitch Glazier, said the music industry was collaborating with responsible AI developers but “unlicensed services like Suno and Udio that claim it’s ‘fair’ to copy an artist’s life’s work and exploit it for their own profit without consent or pay set back the promise of genuinely innovative AI for us all”.
The Big Copyright Question
According to copyright lawyer Kalindhi Bhatia, the defense of Udio and Suno that their use of music for training AI and its subsequent use falls within the contours of “fair use” is thwarted.
"The purpose of use of their AI-based music generation can be commercial, as opposed to merely creative. Further, the volume of ‘use’ enabled by the AI platforms is too high to fall within the ambit of fair use," she said.
Bhatia explained how the rights in the music industry are relatively more complex compared to other art forms, attributing this to "the nature of rights involved and the relationship dynamics between the artist, distributors, record labels, and external stakeholders like streaming platforms and film production houses that are licensed to use the music."
Music rights are typically divided into publishing rights (for the composition and lyrics) and master sound recording rights (for the recorded version). Bhatia detailed the three kinds of royalties: performance (when a song plays publicly), mechanical (for reproducing songs), and sync licensing (for use in movies or ads).
"However, due to the complexity and lack of transparency in royalty distribution, rights often centralise with record labels or distributors. With music, due to the potential forms of dynamic use, rights are generally concentrated with the one entity that yields more leverage," she said.
Indian Laws on AI-Generated Artwork and Copyright Ownership
The advent of AI-generated artwork has raised substantial concerns of infringement. Copyright lawyer Rohin Dubey explained how easy it is for bots to access copyright-protected data and create human-like expressions of art.
"The data is plundered from everywhere: digital libraries, internet databases, and social-media feeds. Those responsible for protecting source material allege that data is siphoned up without any consent or credit," he said.
Speaking in the context of Indian law, Dubey pointed out that AI programmers argue for a safe harbor in the ‘fair dealing’ doctrine of the Indian Copyright Act, which can potentially protect AI training without licenses. "Under Section 52 of the Act, data mining is permissible for certain purposes, one of which includes 'the reproduction of any work by a teacher… in the course of instruction'," he said.
Dubey highlighted their stance, insisting that “access to a wide array of training sets is fundamental to improving AI, without which, the industry may die before it truly starts.”
Explaining this perspective, Dubey noted that art is never created in isolation but is inspired by the knowledge and influences that artists have absorbed. Therefore, any new artwork can be seen as building upon or referencing previous art.
"Do we draw a distinction between a human training their skills on copyrighted music and then writing songs based on music they have learned and a machine doing the same thing?" Dubey asked.
But what does Indian law say about the ownership of a piece of music created by an individual using an AI tool such as Suno AI?
According to Dubey, in such a scenario, the law may either deny copyright protection entirely or attribute authorship to the creator of the machine. Under Section 2(d)(vi) of the Indian Copyright Act, an ‘author’ means “in relation to any literary, dramatic, musical, or artistic work which is computer-generated, the person who causes the work to be created”.
Dubey finds this definition problematic as it lends to the confusion of whether the law recognises the creator of the program or its user as the individual who ‘causes the work to be created’. "Put differently, should copyrights be conferred on the maker of the pen or the writer? This is a difficult question to address unless we find a solution," he said.