Copyright in the age of machine learning

It would be advisable for the Nepal’s parliament to consider incorporating machine learning into their policymaking process including crafting a modern copyright laws to remain competitive in the 21st century.

Bimal Pratap Shah

  • Read Time 5 min.

On April 26, World Intellectual Property Day was celebrated around the world to emphasize the crucial role of patents, copyrights, trademarks, and designs in shaping modern life and also to encourage global innovation and creativity. This year, one of the foremost challenges has been to establish a comprehensive legal framework that recognizes the creative potential of machines, as most countries’ copyright laws do not currently acknowledge AI-generated creations and innovations. The emerging possibility of AI owning intellectual property has brought to light the need for updating current intellectual property laws to keep pace with technological advancements. It’s crucial to initiate a debate on this matter in Nepal’s parliament, given that the assembly has yet to address the subject.

Recently, a track generated by AI featuring imitated vocals of Drake went viral sparking conversations about the capabilities of AI in music production and the potential implications for the future of the entertainment industry. The use of AI-generated music could potentially revolutionize the industry, allowing for faster and more efficient production, as well as the creation of entirely new sounds and genres. It appears that no profession is immune to the disruptive effects of AI technology including those of musicians.

Following the viral success of Drake’s AI-generated music, pop singer Grimes, who happens to be the mother of Elon Musk’s child, wasted no time in declaring that she would split royalties 50/50 with any musician who utilizes AI to replicate her voice and produce original tracks. Her announcement comes amidst a surge in the use of AI-generated music featuring the voices of popular artists. Grimes is not new to the world of AI as she has previously collaborated with the mood music startup Endel to create an AI-generated lullaby app for her son. She has also speculated that if AI achieves mastery in both science and art, it could signal “the demise of human art.”

Joe Russo, who co-directed blockbuster movies such as “Avengers: Endgame” and “Avengers: Infinity War,” revealed that AI could be creating full-length feature films in two years. AI is already capable of making movies, editing actors, and faking voices. While the use of AI in filmmaking has garnered significant attention, media and legal experts have raised several concerns over its implementation, given the murky legal landscape surrounding this emerging field. Even Hollywood, which is at the forefront of technological advancements, seems to be feeling the impact AI’s in filmmaking.

Traditionally, releasing movies in multiple languages has been a costly and labor-intensive endeavor. However, with the aid of AI, it’s now feasible to map actors’ faces and fine-tune their lip movements with greater precision, enhancing their alignment with different languages. This advanced technology, known as ‘Visual Dubbing’, is particularly advantageous for movies released in foreign countries where the original language is not spoken. Visual Dubbing can expedite the process of AI reshoots while ensuring that the lip movements are in perfect harmony with the translated visuals. Thanks to AI, it’s now possible to release movies in every language spoken worldwide without incurring exorbitant costs.

In the near future, it will be possible to train machine learning models on the distinct movie-making styles of celebrated directors, such as Christopher Nolan, Darren Aronofsky, Anurag Kashyap, and Kiran Rao. This will lead to the creation of films based on a short storyline formulated by the audience. Machine learning is a subfield of artificial intelligence that involves developing algorithms and models that enable computers to learn and improve their performance on a specific task or set of tasks without being explicitly programmed to do so. With such training, it will soon be possible to generate thousands of new movies that adhere to the distinct styles of these celebrated directors, allowing one to pick their favorite actors or even AI generated actors in mere moments by using prompts that reflect the style of ChatGPT.

Governments can also use AI to enhance their services and operations.  Currently, the parliament’s policy-making process is slow, and it takes a long time for them to craft, discuss, and finally pass policies. This is because policymaking is inherently complex, bureaucratic, and often subject to political influence. Gathering evidence to support proposed policies or to evaluate their effectiveness can be a time-consuming process. Furthermore, policies tend to be focused on specific areas, such as health or education, rather than addressing the broader socioeconomic issues they are meant to tackle. As a result, inefficiencies, redundancies, and inconsistencies are common.  AI can offer a faster, more comprehensive, and more rigorous approach to policymaking in the short term and help deliver on the promise of a more responsive and inclusive government of the future. To ensure that policies and laws remain relevant and effective in the 21st century, it would be advisable for the Nepal’s parliament to consider incorporating machine learning into their policymaking process including crafting a modern copyright laws to remain competitive in the 21st century.  

There is no single artist responsible for a single AI art piece as AI art is created by algorithms from cross-wired information gathered over time. As result, it will be difficult to use traditional copyright standards to determine who should hold the copyright for AI art pieces. Governments in many parts of the world are still clueless, but the US government has acknowledged the challenge and is working towards a solution. The US Copyright Office (USCO) has recently published a policy stating that an AI-generated work may be eligible for copyright protection if a human has put a substantial amount of creative effort into its final output. At present, the US law states that machines and generative AI algorithms cannot be considered authors, and their outputs cannot be copyrighted. The current copyright laws do not provide protection for works created solely by AI tools like DALL-E, Stable Diffusion, Midjourney, ChatGPT, or the newly launched GPT-4, which include digital art, poems, and books.

The current copyright laws do not provide protection for works created solely by AI tools like DALL-E, Stable Diffusion, Midjourney, ChatGPT, or the newly launched GPT-4, which include digital art, poems, and books.

An encouraging development in the realm of copyright protection has occurred with the registration of a copyright certificate in the United States for a graphic novel featuring images generated by the AI tool Midjourney. The overall composition and words were protected by copyright since they were selected and arranged by a human, but the individual images themselves were not. The United States Copyright Office (USCO) has recommended that those seeking copyright protection for material produced using AI explicitly disclose how the software was employed to generate the content and specify which aspects of the work were created by humans. Failing to accurately disclose such information or attempting to conceal the fact that the material was generated using AI could result in the revocation of one’s certificate of registration and lead to a loss of copyright protection for their work.

It appears that the Nepali parliament may not have fully acknowledged the influence of AI on the international landscape, including the entertainment sector, as the topic has not been broached in the assembly. Merely celebrating World Intellectual Property Day with fanfare is not enough. The parliament needs to discuss issues on AI it takes action to not only update their copyright laws, but also establish or revise entertainment-related laws to ensure the protection of Nepali artists and the entertainment industry. By doing so, the parliament can address the unique challenges posed by the advent of new technologies like AI, while simultaneously promoting creativity and progress in the Nepali entertainment sector as well as maintaining a balance between protecting intellectual property and allowing room for creativity and advancement.

It’s dismaying that the parliament seems to be neglecting the concerns of modern economy in favor of many outdated topics. Therefore, it is crucial to elect highly educated and capable politicians who are well-versed in global policy and economic trend.