Who Holds Copyright Over AI-Generated Art? Justice Gautam Patel Sheds Light on Emerging Legal Challenges.

Who Holds Copyright Over AI-Generated Art? Justice Gautam Patel Sheds Light on Emerging Legal Challenges.

Justice Gautam Patel, a former Bombay High Court judge, emphasized the urgent need for legislation to regulate ‘deepfakes’ and flagged the potential challenges in addressing their implications. Speaking at LiveLaw’s 5th Shamnad Basheer Memorial Lecture on AI, Trademarks, and Copyright: Emerging Issues, Justice Patel remarked, “The challenges ahead involve introducing new legislation and raising awareness about detecting and recognizing deepfakes.”

Justice Patel elaborated on the concept of deepfakes, explaining how AI-generated fake images and videos can convincingly deceive viewers. He highlighted the role of generative adversarial neural networks (GANs) in creating deepfakes, describing their process: “One machine generates a fake and submits it to another machine for evaluation. The second machine identifies flaws and feeds them back to the first, which iteratively improves the fake until it becomes almost indistinguishable from the original.” He compared this adversarial process to the back-and-forth seen in courtrooms.

Highlighting the grave implications, Justice Patel observed, “When AI-driven deepfakes become so sophisticated that even brand owners or judges cannot discern the fake, it creates enormous challenges for protecting trademarks and brand reputations. Imagine the evidentiary burden in court when dealing with an indistinguishable AI-generated deepfake—it’s enough to keep IP practitioners awake at night.”

Copyright Law and AI: A Gray Area

Justice Patel also addressed the complexities surrounding copyright law and AI, citing the ongoing lawsuit by news agency ANI against OpenAI in the Delhi High Court. The case raises a pressing question: Who owns the copyright in works generated by AI— the user who prompts the AI or the AI itself?

He noted that AI-generated content may lack the human effort traditionally required for copyright protection under the ‘sweat of the brow’ principle, which grants copyright to works reflecting considerable effort and skill. He pointed out that AI is not merely a tool assisting creativity but, in some instances, acts as a creator itself.

He illustrated his point by referring to AI-created works, such as a computer-generated painting mimicking Rembrandt’s style or a short novel written by a Japanese computer program that advanced to the second round of a national literary competition. Justice Patel also mentioned DeepMind’s AI, capable of composing music. “This newest iteration of AI is not just a tool—it is a creator, making decisions and generating original works,” he remarked.

Justice Patel shared an experiment where he prompted ChatGPT to produce opening lines mimicking Leo Tolstoy’s style. While the AI generated the text, he noted his role in crafting the prompts and questioned whether his unique arrangement of keywords might itself qualify for copyright. The same challenge extends to AI-generated images, which are often created using specific textual prompts. “In such scenarios, who holds the copyright?” he wondered, describing the current state of copyright law as a “twilight zone.”

Trademark Challenges in the Age of AI

Justice Patel also explored the implications of AI for trademark law, particularly with AI’s ability to analyze patterns and create seemingly original works. He noted that AI’s pattern recognition could put registered trademarks at risk of losing their uniqueness, as AI could quickly generate similar marks or logos that might compete for trademark protection.

“Trademark distinctiveness is critical for protection, but AI’s generative power can undermine this distinctiveness by creating new materials capable of trademark protection,” he observed. He further questioned whether AI-generated trademarks could meet the legal threshold for distinctiveness.

Justice Patel proposed that AI could be harnessed as a tool to counter trademark infringements. For instance, AI could monitor online platforms and social media to track trademark violations and detect deepfakes. He suggested high-value brands could leverage AI for tasks such as generating proprietary variations of brand names and logos to stay ahead of potential infringements.

Concluding Thoughts

Justice Patel left the audience with a thought-provoking question: “Is this entire address of mine original or wholly AI-generated?”

Comments

No comments yet. Why don’t you start the discussion?

Leave a Reply

Your email address will not be published. Required fields are marked *