The Express Pharma first published this article

Increasing usage of AI in every sector 

The main reason for the increasing popularity of Artificial intelligence (AI) is that it develops programs or models resembling human deduction or reasoning. A subset of AI known as “machine learning” uses algorithms trained on vast amounts of data to enable the computer to learn with ever better accuracy without being explicitly programme.  

Massive volumes of data are generated in the biopharmaceutical and healthcare industries, including information on the properties and traits of medicinal molecules, biological, genetic, and clinical data, the effectiveness of treatments, adverse events and dangers, and electronic health records. Numerous sources, both public and private, may provide the data. 

AI systems trained on such data can increase productivity and save costs by streamlining and optimising drug development, including drug discovery, disease diagnosis, risk identification, planning clinical trials, and forecasting safety and efficacy profiles. 

AI can utilise algorithms to analyse massive datasets about various chemical substances and suggest new therapeutic candidates for future testing. Instead of using conventional medicinal chemistry techniques, AI has the potential to “conceive” new chemical entities with expected efficiency and safety properties by modifying current chemical compounds using algorithms and biological and chemical data. AI can also discover novel applications for already-existing medications or chemical substances. AI may also analyse past clinical trial data to forecast future outcomes and dangers of medication candidates. 

AI may analyse medical histories, genetic information, and health records to discover candidates for clinical trials. This improves efficiency and recruitment. Lastly, personalised medicines based on genetic information and particular medical histories may be developed by AI employing algorithms and patient data. 

In other words, AI has the potential to transform the pharmaceutical industry completely. What are the patent-related difficulties that must be taken into account while employing AI to produce medications? Can AI-generated medicinal molecules, for instance, be patented? Who owns inventions made by AI? Some of these difficulties will be covered in this essay. 

 What will be the patent eligibility? 

India has shown an impressive improvement in the pharmaceutical industry in the last 4 years. The booming trend of IP and patent registration in India ascertained that from $3.2 billion in 2021 to $1.4 billion in 2022, India’s overall investment in health technology startups fell by 55%. Before 2019, India’s healthcare industry received US$ 586.93 million in total investment. Therefore, it can be said that the Indian healthcare business is growing at a good rate despite the COVID-19 situation acting as a hindrance.   

It is anticipated that India’s medical device industry will be worth $50 billion by 2025. Due to the COVID-19 problem, India has a unique chance to alter some of the rules controlling the distribution of medications and drug testing.  

Considering the practical scenario of the Patent process in India, explaining how an AI model functions to turn the assertion into a creative idea could be challenging. However, a candidate can continue to concentrate on the details of the data input and the generative output of an AI model. 

Let’s have a glance at the eligibility criteria for patents in India: 

• Novelty: To determine a creation’s patentability, wonder is essential. The design must produce new knowledge, a new item, or a new process. Any document, patent granted, published patent, non-patent literature, or other work already in the public domain should not be used to anticipate it. It must be different from what is already understood. 
• Inventive Step: The inventor must make a creative contribution to the invention. It needs to be something that a skilled craftsperson wouldn’t anticipate. Let’s say an innovator creates a device to address a technological issue. A different expert in the same sector offers the same solution by drawing on his knowledge or absorbing instruction. The inventor’s technical solution will only be considered original if it was a suggestion or motive. 
• Industrial Application Capabilities: Section 2(ac) of the Patents Act states, “the creation is potent of being utilised or manufactured in a sector.” It implies that a product must be helpful and patentable because the invention cannot exist in the abstract and must apply to all fields. 
• Specification: The inventor must submit a patent application with a specification to get a patent (Section 10). The objective of the specification is to provide thorough information to the public about the innovation and the means of carrying out and defining its scope.  

When discussing the eligibility of artificial intelligence as an inventor, we need to explore the scope of inventors. 

 Can AI make itself eligible as an inventor? 

Before we dive deep into whether AI can be considered a valid inventor, we need to understand the concept of AI utilised in invention works. Artificial intelligence is a helping hand in several industries, including improving processes, enhancing industry productivity, and impacting operations and revenue. 

Artificial intelligence’s increasing ability to devise innovative solutions is throwing a challenge to the boundaries of the existing patent laws that were beyond the imagination of the legislature. The new draft of the Patent law in India has brought several necessary changes to be implemented for better patent regulation. However, it still does not consider AI as a valid inventor. 

Taking the reference from the US patent law, it is worth mentioning that the USPTO considered that question in 2019, publishing a request for comments on patenting artificial intelligence inventions. The query involved whether current patent laws adequately address inventorship for AI-created innovations, typically created without human intervention. 

Many commenters agreed that new policies are required to ensure that AI-generated inventions are adequately recognised and protected, even though the responses to the USPTO’s request showed a wide range of opinions on AI patent inventorship. 

This case added a valuable opinion to help the USPTO as it investigated the ethical and legal concerns related to AI inventorship.  

The case involved two patent applications submitted in 2019 by Stephen Thaler, the creator of the AI system DABUS (Device for the Autonomous Bootstrapping of Unified Sentience)— who was listed as the inventor. 

The USPTO denied the applications because there was no legitimate inventor, deeming them deficient. Using U.S. Court of Appeals for the Federal Circuit rulings that reached identical results in denying inventorship status to states and companies, the USPTO held that the Patent Act only grants inventorship to natural persons. 

In another case, Thaler vs. Vidal, the Federal Court ruled that the inventor must be a natural person. AI is not a natural person; an invention solely made by AI would not be eligible for patent protection or the title of inventor. 

 Conclusion 

By following this line of reasoning, AI would be viewed as a tool humans use to carry out scientific study, much like a person may use a computer, a DNA sequencer, or other laboratory apparatus. The person employing the tools would be given credit for the finished product. As a result, if AI creates a chemical compound with the potential for activity, the idea for the compound would be credited to the person who oversaw or guided the AI to produce the compound’s structure. 

Furthermore, it may be argued that an innovation is not a “permanent idea of a complete and operative invention” until a human mind affirms its usefulness. No matter how independent artificial intelligence becomes, human intelligence must still be verified.