We examine the ownership of AI-generated patents and describe recent examples of the friction between AI and IP and potential policy changes to address these challenges.
What is AI?
Before diving into the intricacies of AI and IP, let’s briefly define AI. AI refers to developing computer systems capable of performing tasks that typically require human intelligence. These systems learn from data, recognize patterns, make decisions, and adapt to new information. This learning system enables AI to perform complex tasks autonomously or with minimal human intervention.
Two Types of AI
There are generally two types of AI. Each of these types of AI may have different impacts on IP rights.
Traditional AI Systems
First, there are traditional AI systems that work from and rely on explicit programming or rules. These rules are typically created by human experts in a specific domain and provide a set of if-then statements that guide the system’s decision-making process.
An example of a traditional AI is a system that helps doctors diagnose a patient based on the patient’s symptoms. The doctor would input all the patient’s symptoms, and the AI would provide a likely diagnosis from the doctor’s inputs. Traditional AI systems excel in domains where expert knowledge can be effectively codified into a set of rules.
Generative AI Systems
The second general category of AI systems is called generative AIs. Generative AI is a subfield of artificial intelligence that focuses on creating computer systems capable of generating original and creative content, such as images, music, text, and even videos.
Unlike traditional AI systems that rely on explicit programming or rules, generative AI leverages machine learning techniques, particularly deep learning, to learn from large datasets and generate new content that resembles the patterns and characteristics of the training data.
Generative AI in the Real World
Generative AI has found applications in various domains, including art, music, storytelling, gaming, and design. It enables the automation of creative tasks, facilitates content creation, and opens up new possibilities for human-machine collaboration.
However, generative AI also raises important ethical and legal considerations, such as copyright infringement, ownership of generated content, and the responsible use of AI-generated outputs.
The remainder of this article will focus on how generative AI technology impacts IP and what can be done to protect IP rights.
Interconnection of IP and AI
AI technology relies heavily on the use and analysis of vast amounts of data. Consequently, many issues surrounding IP and AI are rooted in the protection and ownership of these data sources and the algorithms, models, and creations produced by AI systems. IP rights, which include copyrights, trademarks, and patents, play a critical role in safeguarding innovation and creativity. But AI poses unique challenges to the existing IP framework.
Issues with AI and Copyrights
Copyrights are designed to protect original creative works fixed in a tangible form, such as literary, musical, and artistic works. Authors of these original works own the right to reproduce, distribute, and publicly display their work for a certain amount of time. With AI, the line between human and machine authorship becomes blurred. AI systems can generate artwork, music, literature, and other creative outputs.
The question arises: who should be granted copyright protection when AI is involved? Should it be the programmer who created the AI, the AI system itself, the human who operated the AI to create the work or a collaborative effort between the three?
AI and Copyright Laws
Currently, copyright laws typically require human authorship, leaving AI-generated creations in a legal gray area. The United States Copyright Office has provided guidance that non-humans are not considered authors who can receive copyright protection.
However, this guidance has made considerations that AI can assist in creating the work if it is merely helping a human who is providing the “traditional elements of authorship.”
This copyright gray area between AI and human authorship is a pressing concern for the U.S. Copyright Office. The number of applications to register AI-generated works will likely increase as access to generative AI increases.
Issues with AI, Trademarks, and Brands
Trademarks serve as identifiers of the source of goods or services, providing exclusive rights to brand owners. With AI-generated content becoming more prevalent, the risk of trademark infringement increases. AI algorithms might inadvertently generate content that infringes on existing trademarks, leading to confusion in the marketplace.
The AI systems could create brand names, logos, or slogans that resemble registered trademarks. Using those generated symbols, an AI could be used to spam or spread misinformation, significantly harming the trademark holder.
What could happen to trademark owners with the influx of AI?
The trademark owners face serious consequences because of the confusion and infringement caused by AI-generated content. The purpose of protecting one’s brand and products with a trademark is to prevent others from producing something identical. Now, with a few instructions from an AI, the hard work of a trademark owner can be easily copied.
However, the problem could also be the solution to these issues. AI systems could be developed to monitor and alert trademark owners of trademark infringement and counterfeits. These systems can be used to promote fair competition and safeguard consumer interests.
Issues with AI and Patents
Patents protect inventions, granting exclusive rights to their creators. When it comes to AI, the question arises: who should own a patent created by an AI system? Traditionally, human inventors are named as the patent holders. However, there has been a direct challenge to this rule where an inventor sought to name an AI as the inventor on a patent application.
One viewpoint is that the organization or individual who owns and controls the AI system should be considered the patent owner. Another perspective suggests granting patents to human programmers or inventors who trained or developed the AI system. Finding the right balance between acknowledging AI’s role in the invention process and preserving the rights of human inventors is crucial. But at this point in the evolution of AI, the AI itself is not considered the inventor of a patent. Under the Patent Act, the inventor of a patent must be a human being.
Recent Examples of AI and IP Friction
In the past few years, there have been many examples of the complex relationship between AI and IP.
DABUS
One example follows from the previous section. Dr. Stephen Thaler attempted to list an AI called “DABUS,” as the inventor on a patent application.
After the United States Patent and Trademark Office denied the application for failure to list a human inventor, Thaler challenged the decision and the law requiring a human inventor. The U.S. Court of Appeal for the Federal Circuit concluded that the Patent Act was clear: “inventors must be natural persons.”
“Heart on My Sleeve”
Another more well-known example comes from the music industry. In April 2023, Drake released a new song featuring The Weeknd, titled “Heart on My Sleeve.” The problem was neither Drake nor The Weekend produced the track.
The record was created by an AI using voice-mimicking technology, and then the song was uploaded to several music streaming sites. The seriousness of this example cannot be overstated. Drake and The Weeknd have both won several Grammy awards, and their music garners billions of downloads. As the popularity of “Heart on My Sleeve” and others produced the same way grows, this type of infringement is not likely to stop.
Policy Changes to Address Issues
To address the complex IP issues arising from AI, policymakers and legal experts must consider several potential policy changes:
- Clarifying Authorship: Amend copyright laws to explicitly recognize AI as a potential author, allowing AI-generated creations to be eligible for copyright protection or not. This could involve creating a new category of authorship that considers AI involvement.
- Evolving Trademark Laws: Update trademark laws to encompass AI-generated content and establish mechanisms to monitor and enforce trademark rights in AI-generated materials, ensuring brand owners’ protection.
- Redefining Patent Ownership: Introduce legislation or guidelines to clarify patent ownership when AI is involved. Consider a hybrid approach that acknowledges the contributions of both AI systems and human inventors, ensuring a fair distribution of rights.
- Standardizing AI Attribution: Develop industry standards and best practices for attributing AI-generated works to their creators. This can help establish transparency and accountability in the use of AI systems while mitigating legal disputes.
- Promoting Collaboration: Encourage collaboration between legal experts, AI researchers, policymakers, and industry stakeholders to navigate the rapidly evolving landscape of AI and IP. This collaboration should establish ethical guidelines and ensure IP frameworks keep pace with technological advancements. This collaboration is essential at this point in AI advancement because it does not appear that AI will suddenly disappear. Because AI is likely to become more constant in everyday life, creating a plan and a dialog between the producers of AI and the lawmakers is crucial.
Conclusion
As AI continues to reshape industries and push the boundaries of human creativity, addressing the intricate interconnection between AI and IP is imperative. By reevaluating existing IP laws, clarifying authorship, evolving trademark regulations, reviewing patent ownership, and promoting collaboration, policymakers can foster a legal framework that encourages innovation while safeguarding the rights of creators and society. Balancing the interests of humans and machines in IP is crucial for AI technology’s equitable and sustainable advancement.
Questions?
If you have questions about how AI may impact your intellectual property rights, contact our experienced IP lawyers today for a free consultation.