Inventions By AI Are Pushing Patent Laws To Breaking Point

Subject matter eligibility issues can arise with computer-based innovations, particularly in the domains of machine learning (ML), bioinformatics, and artificial intelligence (AI). The United States Patent and Trademark Office (USPTO) may receive objections about subject matter eligibility that might prevent a patent application from being approved and potentially result in the post-grant invalidation of a patent. Identifying the computer-related invention as an abstract idea, which includes mental processes (concepts that can be performed on paper and a pen), methods of organising human activity (such as managing interactions between people), and mathematical concepts, is one of the subject matter eligibility challenges.   

In exchange for their respective discoveries, inventors who hold patent rights also have the right to prevent others from using (or otherwise violating) those inventions. According to US patent law, someone who “makes, uses, proposes to sell, or sells any patented invention inside the United States or imports any patented invention into the United States within the period of the patent therefor” is infringing on a patent claim.  

 A two-step investigation is necessary to establish infringement to: (1) ascertain the meaning of each phrase in a patent claim; and (2) demonstrate that the accused device satisfies each claim term (i.e., claim limitation), either explicitly or in accordance with the theory of equivalents. 

 Given that most AIs now have the technological ability to violate patent claims, liability in circumstances where AI is the infringer of patent rights is another crucial patent law problem that will probably be disrupted by AI.  

The liability problem presents the question of who should be held accountable for acts made by AI – the end user, the developer, or AI itself- as well as the associated question of how to measure culpability. 

Solution can be: 

• As stated in the Resolution, the “laws regulating culpability for harmful activities – where the user of a product is accountable for a behaviour that leads to harm” may apply to damages brought on by AI. One potential defendant would be the AI’s end users. However, this could cause customers to become more concerned and less eager to adopt potentially helpful AI. In many cases, especially when they are people rather than smart organisations, end users are unable to anticipate the patent violation. 
• This brings up the alternative of holding the AI developer or manufacturer responsible. In patent litigation, it is typical practise to hold the maker of a product accountable for patent infringement. This may be appropriate in the context of artificial intelligence as well because the developers are ultimately responsible for creating the AI (that violates the patent), are typically in a better position to anticipate the infringement than the end users, and have probably reaped financial benefits from the AI. 
• However, can a human agent genuinely foresee against or adequately supervise the AI to prevent violation in the case of totally autonomous AI? If humans were held accountable for unforeseen behaviours, such as violation of patents, would this discourage the development and usage of artificial intelligence (AI)? If so, how would this affect innovation?  

The legal system will need to change as more businesses and inventors employ AI to produce new discoveries. The standard for obviousness in the patent process will rise once patent examiners are forced to assume that a PHOSITA has access to AI. It’s not like that anymore. Currently, AI offers a significant competitive edge, and businesses creating new AI processes or utilising it to create new ideas may be sitting on a gold mine. 

In case of AI as an inventor, a long discussion is required to discuss about how to define a POSITA i.e., Person of (ordinary) skill in the art. 

• There needs to be debate over whether the current definition of a POSITA, which calls for a human rather than an automaton, is sufficient or whether it has to be changed to include the possibility of an AI-equipped person if the usage of AI is widespread in that field of technology. 

The primary goal of patents was to promote innovative technologies, scientific advancement, and industrial growth. According to patent law, the inventor has a monopoly on using their patented goods, albeit they may provide permission in advance and in exchange for a fee for someone else to use it. The question of whether AI-generated innovations should be granted patents must be considered in the context of whether doing so will advance the basic goals of the patent law system.  

• One point of view is that extending patent rights to ideas created by artificial intelligence (AI) will speed up innovation and maybe even allow for advancements that would not have been achievable with only human creativity. Others have suggested that patent protection does not encourage innovation, regardless of whether ideas are produced by humans or artificial intelligence (AI). According to this theory, more patents issued as a result of inventions produced by AI would hinder new enterprise entrance, raise social expenses, and create monopolies. This will inhibit innovation. 
• On the other hand, even if patenting AI-generated ideas eventually helps innovation, doing so may negatively affect future human creativity as the replacement of human invention with autonomous algorithms might result in the decline of human intelligence,” according to a study. Reduced creative talent raises concerns that it may result in the loss of excellent research and development (R&D) jobs or perhaps whole R&D-intensive enterprises.