OPINION | Can you claim ownership of your AI-generated creatives?

Today we are being driven by Gen AI-led innovations. Such innovations dominate the Indian start-up space, with a substantial percent of the patents granted in the last decade belonging to innovations that have AI-related applications.

Transport and healthcare sectors have been engaged in creating top-notch patent worthy AI applications. From efficient traffic management dashcams for monitoring congestions and detecting traffic violations to making wearable devices for monitoring foetal health of pregnant mothers – these patented AI innovations have transformed civic and private lives.

Copyright laws lag Gen AI’s impact

And yet, the copyrights laws have not kept pace with the acceleration seen in the AI ecosystem. The existing legislation struggles with answers when Gen AI enters the picture with creative capabilities, particularly when assessing two key conditions – determination of originality and authorship.

To recognise AI-generated work as copyrightable is a complex assessment, exemplified in the intriguing case of Ankit Sahani.

Inconsistency in cross-border application of copyright law

Sahani commissioned an AI tool Raghav AI to blend a photograph of a sunset that he had taken with Vincent Van Gogh’s masterpiece Starry Night. He titled the resultant image SURYAST and named both himself and Raghav AI as co-authors in the copyright applications filed in India and the US. The applications were denied in both the jurisdictions on the principle that copyright protection recognises only human authorship, which was absent in the AI resultant image.

By contrast, Sahani secured copyright registration of his image in Canada.

Berne Convention is now on shaky ground

This divergence creates regulatory dilemma in India, which is amplified when taking into account the Berne Convention - an international treaty for copyright protection. Signed by 180-member countries, including India, US and Canada, this treaty says copyright registered in one country is enforceable in others. Which begs the question: given its Canadian copyright registration, will Indian IP law be enforceable if SURYAST sees infringement in India and ignore the mandatory originality and authorship conditions?

Grant of selective copyright in Kashtanova’s case

In another case, a copyright granted to artist Kris Kashtanova’s comic book Zarya of the Dawn was revoked by the US Copyright Office when it became known that Kashtanova had used AI tool Midjourney to create the images.

Initially, Kashtanova was listed as the sole author of the comic book as Midjourney’s involvement in creating the images was not divulged. However, Kashtanova argued before the Copyright Office that the creation of each image was guided by human prompts to the AI tool, and therefore, the images involved human inventorship or human authorship. But the Copyright Office, after reviewing the images, granted only a selective copyright – for the arrangement of text and images.

Together, these cases reveal discrepancies and gaps in intellectual property protection in the context of AI-supported creations. Artificial intelligence is not only disrupting the creative world, but also emerging as a creative contender itself. Because of the scale of innovations taking place in our backyard, there is a need for clarity of the law in protecting non-human inventions or AI-generated outputs guided by humans. We wait for answers to the looming question of authorship when the creator is not human.

 (Swati Sharma is Partner (Head- Intellectual Property), Cyril Amarchand Mangaldas.)

Views are personal and do not represent the stand of this publication.