Will AI disrupt the IP regime? Human Intelligence vs. Artificial Intelligence
- What is Artificial Intelligence?
- What is Intellectual Property?
- Types of AI Systems
- AI disrupt the IP regime: Explain
- AI and Trademarks
- Impact of AI on Trademark laws
- AI and Copyright
- Impact of AI on Copyright Laws
- AI and Patents
- Impact of AI on patent laws
What is Artificial Intelligence?
There is no single definition of AI that can be considered proper or acceptable; however, from my understanding of AI, I can define it as; an algorithm that mimics human intelligence by use of machines, particularly computer-generated works and other related technologies.
Artificial Intelligence (AI) is slowly trying to replace Human Intelligence (HI) to the extent that it is now performing most of the automatic complex intellectual tasks. Consequently, AI is trying to outperform human capabilities. However, it fails to link human intelligence and consciousness.
AI’s capabilities extend towards building a neural network technology designed to work closely identical to how the human brain functions. However, the absence of moral and legal rights in AI systems limits AI’s ability to own property, as most objections to artificial inventor projects focus more on AI ownership. These aspects of AI are creating ripples in the legal framework, particularly Intellectual Property Laws and regimes. Thus it would be a milestone if IP laws could be reviewed to create new legal frameworks, particularly on the aspect of legal persons and grant them rights to factor AI tools and AI-generated works in various IP laws.
What is Intellectual Property?
Intellectual property refers to the creation of the mind, including artistic works, inventions, designs and symbols. Intellectual Property laws such as Patent laws, Copyright laws and Trademark laws are put in place to protect various IP rights such as patents, trademarks and copyright. However, With the rapid growth of AI, some major IP rights are greatly being disrupted by AI, including; the patent law and patent protection of AI innovation; Copyrighted work; and trademark laws, which have formerly withstood the E-commerce and social media era and now the new challenge is AI.
Types of AI Systems
Autonomous Intelligent Systems
This is an AI software system that acts independently of direct human supervision, i.e. the problem is identified by a human creator while an AI machine outlines the process and comes up with a solution or decision, e.g. self-driving cars and smart manufacturing robots. According to a European Commission report on AI and IP, it declared that we are slowly moving towards AI autonomy. However, from these definitions, an AI system cannot be considered fully autonomous. This is because, in a typical machine learning system, human creators, operators, designers, or system administrators are still accountable for its process.
Assisted Intelligence System
This AI system automates simple processes and tasks by connecting the joint power of Big Data, cloud and data science to aid in human decision-making, i.e. machines do the action while human makes decisions. Basically, this system improves what is already being done by humans or organizations.
This system focuses on the technology’s assistive role, i.e. machines do the action while humans collaborate with machines in the decision-making process such that humans can do more than their capabilities.
AI disrupt the IP regime: Explain
AI and Trademarks
A trademark is a type of Intellectual Property right consisting of a sign capable of distinguishing the goods or services of one business from the other. The way products and services are purchased have revolutionary changed over the years, and the same has affected requisite changes in the development of Trademark law and regimes. The various changes in how products are purchased have caused a significant change to the trademark laws. For example, the process of purchasing products has evolved as follows:
a) In the 19th and early 20th centuries, people used to buy products with a third-party shop assistant who provided a filter between the unbranded product and the consumer by advising the consumer on the products based on their knowledge of the product.
b) This later evolved with the advent of modern supermarkets, and the decision to purchase the products shifted entirely to the consumer as they could access all the products. The consumer’s product choice in modern supermarkets is largely influenced by product branding. Thus, brands replaced third-party shop assistants. During this process, the consumers developed an emotional bond with the brands to the extent that they interacted with each other.
c) In the 21st century, purchasing products evolved once more, particularly with the introduction of e-commerce and social media, which allowed consumers to gain more knowledge on products and more products available to consumers. Equally, the consumer’s product choice is dependent entirely on themselves as they have complete control over the purchase decision. However, social media largely influences consumers by using the ‘like’ button, influencers and celebrities, etc., who promote product brands.
Now, the structure of the purchasing process has changed, and the average consumer has been replaced by AI. Consumers now prefer online shopping through websites such as Amazon website (or modern shop assistant), which is an AI application. This website recommends products to consumers based on their browsing and purchasing history. Thus, AI application becomes a filter between the product and the consumer and the emotional bond is ruined and replaced by an artificial bond.
Generally, AI systems suggest products based on the price and speed of delivery, and these criteria are causing great damage to both the brand and the consumer, making it a struggle for trademark laws to accommodate these AI-related emerging issues.
We also have AI products such as Alexa of Amazon and Siri of Apple, which can interact with humans. Alex has the capability of ordering products online based on available brand information. However, the question lies as to what happens when only selected brands are inserted in the AI system and on what basis does Alexa suggest the brands? All these AI systems can promote issues such as unfair competition.
Impact of AI on Trademark laws
Since an artificial consumer is replacing the average natural consumer, then the fundamental features of the trademark law are bound to hit. Trademark laws were intended to facilitate the purchase process by providing a platform between the natural consumer and the brand, i.e. human has faults, and trademark laws were meant to fill these faults. AI is therefore challenging the natural consumer with its perfect characteristics. The foundation of trademark law is found on terms such as “imperfect recollection”, “secondary infringement”, “average consumer” and “likelihood of confusion”, which need to be reconsidered, especially with this rapid technological advancement.
Equally, the common factor in trademark infringement is the liability issue; therefore, if an AI product such as Alexa orders a product without human interaction, who will be the average consumer? Likewise, who will be liable for trademark infringement for the purchase made by an AI system?
AI and Copyright
Generally, copyright is associated with the human creative spirit, which recognizes human input in artistry and encourages the expression of human creativity. Robotic artists have been involved in various types of creative works. For example, Dr. Eldon Tyrell (Blade Runner) created Rachel (a replica), and various issues arise here, such as whether Dr. Tyrell can enjoy copyright protection in Rachel and her output. Considering that Rachel’s works relied greatly on the creative input of Dr. Tyrell. Another example is where AI generates creative works such as songwriting and gaming, which can be deemed free from copyright since they are not created by humans.
What will happen to the companies that have generated the AI that has created songs and games, knowing that they have invested millions in such systems and the same is not protected by law?
Another issue is if copyright could be assigned to AI-generated copyright works, then in whom should the copyright vest? If the answer to this is affirmative, then should consideration be given according to the legal personality of an AI application, particularly where the work is created autonomously, so that the copyright would vest in the legal personality and the personality be governed like in an organization?
Currently, copyright laws handle such works by either entirely denying copyright protection or attributing authorship to the creator of the system.
Impact of AI on Copyright Laws
For creative works to qualify for copyright protection, there has to be an originality aspect which ideally requires a human author. Therefore, computer-generated works based on AI technology will greatly impact copyright law because, traditionally, it was not an issue of concern as the system was only an AI tool that supported the creative process.
In modern creativity, the issue is whether copyright law should recognize the contributions of the programmer (human intelligence) or the program user (AI). This is now the problem in the digital world.
With the rapid growth of AI, this system is used more than just as an AI tool as it is more involved in AI inventorship aspects such as decision-making in creativity with little or no human intervention.
Another issue in copyright laws is the duration of protection of the copyrighted works, which in most jurisdictions is 70 years after the author’s life. However, in AI-generated works, this will not make sense for AI because the author is not bound to die and not be alive in the first place.
AI and Patents
A patent is an exclusive right or a document that a government grants to an inventor for an invention, such as a product or process that provides a new way of doing something or offers a new technical solution to a problem.
An inventor is someone who engages in creative activities and identifies themselves as such in a patent document. In most jurisdictions, an AI system cannot be named an inventor, and only a natural person can be recognized as an inventor. This is because most of their Patent laws stipulate that in a patent application, the patent documents must specify both the inventor (name and address) and the invention, making the scope of AI very restrictive as AI inventions lack these requirements.
A good example of this is the Patents Act of the United States, which defines an inventor as “the individual” or “individuals (collectively)” who discovered the subject matter of inventions. Since its Patent law does not define “individual,” the Federal Circuit, in a Supreme Court decision, held that an “individual” is understood to be a human being. Thus only a natural person can be an inventor. Similarly, the UK law, the Federal Court, the European Patent Office and German Federal Patent Court declined to grant patent applications with no human inventor.
Previously, Artificial Intelligence played an assistance role in invention processes, but with the development of AI innovation, the role of AI inventorship and creation has shifted to independent creations. Now, the question arises as to who can be granted patent protection for AI inventions.
Impact of AI on patent laws
It is now becoming tough for the courts and other systems worldwide to establish jurisprudence on ownership of AI, particularly when more patent applications identify an AI system as an inventor. This is because most patent laws are based on the perception that inventions are based on human intelligence, making humans the rights holder of the innovation as opposed to AI-generated works.
However, in some jurisdictions such as Australia, an AI innovation can be named as an inventor in a patent application based on a certain Australian Federal Court decision which relied on the Australian Patents Act, held that an AI system could be named as an inventor on a patent application. Therefore, if courts and other systems can decide to consider AI-generated inventions to be patented, then such decisions will greatly impact patent laws.
To consider AI inventions as inventors or owners on a patent application, Patent laws can adopt either one of the following approaches; –
- Treating Artificial Intelligence as a legal person by establishing its legal personality that is subject to rights and obligations thereto; or
- Granting patents to Artificial Intelligence without specifying any inventor.
While you put your heart and soul into creating something, you should also be aware of the legalities, how the law will treat your invention, as well as the rights you’ll have as an inventor. In order to protect your invention, you should speak with an experienced intellectual property lawyer once your invention is ready for public release.
Another issue relates to what happens if AI infringes a patent. Most patent laws recognize a human as a patent infringer but have no provisions for non-human infringers. For example, who will be liable when an AI application inadvertently duplicates a patented invention? In such an instance, will the system have to lift AI’s veil and direct liability to the person who created AI, i.e. manufacturer or operator?
AI is evolving rapidly, which will undoubtedly affect every facet of the IP system. Therefore, there is an urgent and constant need to examine and revamp various IP laws so that they do not merely become a repository of loopholes that cannot sustain various IP rights. In addition, the rapid growth of technology is changing the way AI inventorship and related technology are involved in the invention process; therefore, a fit-for-purpose IP law is needed to ensure it serves both computer-generated work and human intelligence.