Computer system is said to have the artificial intelligence (AI) when it is able to replicate and mimic the human cognition functions like learning, reasoning, problem-solving, and decision making. The disruptive nature of AI can be seen in various industries, such as healthcare, finance, and transportation. With the proliferation of AI-based inventions, an important question emerges, namely, can artificial intelligence be patented? The patent law permits inventors to enjoy rights to their inventions so that they do not allow other people to replicate, use or sell them within a given period. Invention to be given patent protection must be novel, non-obvious and must be applicable to industry.

This paper is a critical evaluation of patentability of AI-based inventions in India, the United Kingdom and the United States. It examines the history of the AI technology, the history of the patent regimes and the legal standards on AI inventions within these jurisdictions. Considering the high development pace of AI, the study highlights the necessity of just and effective intellectual property systems that would promote innovation without causing a lack of legal clarity. Moreover, it underlines the ongoing problems of determining the novelty and the innovative step in AI-produced work. The results highlight the necessity of creating guidelines and universal assessment criteria that could be used to define whether AI is patentable, and encourage the further investigation of the role of AI in the innovation process and in intellectual property law