• The Law Gazette

Implementation of Artificial Intelligence in the Realm of IP Rights: Need of the Hour?

The essence of a developed country is determined by its growth and development. In a country like India, there has been an enormous boost in technology which has turned out to be a great asset in terms of our economy. So far, every advancement has proved to be beneficial by bringing in more scope, one such being, Artificial Intelligence(“AI”). The transfigure of wealth from tangible control to intangible control like patents and copyrights have portrayed international recognition and development in the field of Intellectual Property Laws. Artificial Intelligence is one the biggest scientific inventions of the century which can change our locus forever. The work that requires a human mind and effort is now overpowered by AI, winged its way in the Indian Legal system.


ARTIFICIAL INTELLIGENCE AND IPR

As a technology, artificial intelligence has proved to be effective and efficient. It has led to innovations in Intellectual Property law mainly with patents and copyrights. Given the current situation of the pandemic, the AI-based systems aid the hospitals in serving food and medicines through Asimov’s KARMI-Bot.[1] The role of artificial intelligence in intellectual property has been a topic of debate and discussion. The use of AI in IPR involves patent protection for AI-based products i.e., RoboReview which conducts patent filings and reviews patents, and drafting. According to Niti Ayog, they believe that artificial intelligence is said to spur the growth rate by 1.3% in 2035.[2] Many social sectors will adopt this method, and interestingly, the global pandemic has fastened its growth. Our country is said to be the front-runners in adopting this technology. The use of a data privacy legal network can protect human rights as well. [3]


The IP laws intended to enable innovations and to produce a reasonable profit for the patent holder. In the financial outlook, it encourages one to come up with new creations which incidentally helps in the economic growth of the country and to export such products on the global scale. Over time, machines are said to be constructed in a way that is adept to learn and to be self-sufficient but, the grant of patent rights to such AI-created machines is challenging compared to a human invention.[4] The question here is, how to bring a balance between human invention and AI-created inventions and how to value them in terms of patentability?


Artificial intelligence and IP overlay each other in various ways. It is a well-known fact that intellectual property rights may be assisted by AI technology where for instance, WIPO translate, and Brand Image Search use the AI application for automated search results.[5] The second overlay of IP and AI is that the former can provide legal protection for the latter as the IP Policy aims to trigger innovative ideas and provide incentives by granting protection to the owners in the economic and technological areas. Patents and Copyrights are specifically suitable for the protection of AI and in certain cases the applicant has named the AI application as the inventor. So, the other uncertainty is whether such applications can be named as the inventor? If so, what are the legal provisions governing this issue? Moreover, the revelation could be difficult for an AI-generated invention as it is dynamic in nature to algorithms of machine learning.[6]


The solutions for these issues may be addressed by the legislator leading to a sui generis system of IP rights that must be made to alter the incentives for AI. The third overlap is the transparency and accountability of IP on AI. For instance, if a particular firm or company gives a low credit score to a person there exists a right to appeal but no such right to demand the algorithms used to ascertain such a poor credit score.[7] This would infringe the rights of the owner of the IP. But the issue is how to suffice the obligation of a machine that involves a constant dynamic development? In this case, a conflict may arise between the rules of IP protection and transparency. However, the revelation here has nothing to do with algorithmic rules, but it is a mere outcome of such an application.


In the field of copyrights law, protection is given to the author for creativity. All the copyrights are concerned with protecting the work of the human intellect. The domain of copyrights is the protection of artistic and literary works. The rights granted under national laws to the owner in a protected work are generally exclusive. In the well-known Monkey Selfie case, it was held that no copyrights can be given to a picture taken by a monkey as it was taken without human intervention and animals are not entitled to registration.[8] Whereas, in a few countries, copyright protection is granted to computer-generated works. The common law authorities have a low origin for creativity for it to be protected. There must exist a direct relationship between the inventor and the invention. In the case of AI, such a connection is absent as behind every AI-generated invention there is a developer who leads to the question of who is the true inventor in this case?


As it is generally known, for a creation to be protected, the rights are to be claimed by a legal person (or company) and there is no fact whether AI has the status of a legal person. The complication arises as to who can be held liable when an AI product infringes the rights of a patented invention. Many such issues require the need to be addressed and worked upon to make use of the technology in its best possible way.


SOLUTION

As we know, nothing fascinating is ever completely one-sided. One of the many disadvantages is that the data which is used for AI algorithms is discriminatory in nature. This undercut equal opportunity leading to oppression. According to a study in 2019, the algorithms used in US hospitals would help to determine which of the patients would need extra care and it turned out that this algorithm favored white patients over the black.[9] In the case of Amazon, these algorithms proved to be biased towards women as in the past few years, it has resulted in hiring more men than women.[10]From this, we conclude that to make non-biased algorithms if the appropriate data used is bias-free and secured, regulated, and implemented by the government, then there is a possibility that our country too can adopt this method for effective and efficient results.


If errors like creativity and non-biased programming are rectified in every facet, the technology of AI can be of great use. There is a pressing need to enforce laws that modulate the inventions created by artificial intelligence machines in all countries to acknowledge the fundamental aspects of producing such machines. There are numerous challenges in the implementation of this technology. The performance of an AI machine should not merely be mechanical but, must prove to be inventive and for this, the first step towards it would be to establish AI as a separate legal entity to acquire the rights equivalent to humans or companies.


Another clash that arises is, it replaces humans with machines which could cause a major setback in the employment of the people. Therefore, they must be executed in a way where there is no loss of jobs. As much as this technology can lead a ‘developing country’ to a ‘developed country’, it tends to reduce the cleverness of an individual to come up with ideas as everything will be replaced by machines soon which can pose a threat to human development. Hence, it needs to be studied carefully keeping in mind all these issues and a proper way to deal with them to achieve the desired results.


CONCLUSION

In various underdeveloped and developing nations, the use of artificial intelligence would be significant. The mere supply of technology is not sufficient in such countries. All the aspects, right from education, humanitarian aid, healthcare and food production must be made more efficient by integrating AI which can be an exceptional change but again, this is far from reality. In a developing country like India, where around 3 crore cases are pending before the courts, the citizens are losing faith in the judiciary.[11] The legal profession is largely competitive in nature as there is more exposure than before and advancement of technology as well which is a huge transformation compared to the last 20 years.[12] As much as AI can help lawyers to find loopholes based on the data uploaded which otherwise would be time-consuming, however, the legal rights and duties are not yet determined for a robot.


Therefore, introducing AI would require a lot of patience and effort and more importantly, there must be clear-cut transparency in the legislations and statutes as to how to deal with the human-machine relationship. These machines also lack creativity as if any command or unrelated information can lead to immaterial results. It would be ignorant to forget that above all, these AI machines are human fed, so the question is, could they think beyond the human mind? Some AI experts like Gary Marcus, Nick Bostrom predict that AI will be able to do anything that humans can but do it better. This is a questionable assumption, but will AI surely surpass humans in specific domains of formal logic, understanding, translation and fault finding?


ENDNOTES [1]https://economictimes.indiatimes.com/small-biz/startups/newsbuzz/how-karmibot-others-are-making-isolation-wards-robot-ready/articleshow/74803309.cms?from=mdr [2]https://analyticsindiamag.com/ai-is-expected-to-boost-indias-annual-growth-rate-by-1-3-by-2035-niti-aayog/#:~:text=According%20to%20a%20draft%20discussion,sectors%2C%20boosting%20AI%20for%20All. [3]https://www.analyticsinsight.net/artificial-intelligence-drives-indian-economy-towards-growth/#:~:text=A%20draft%20paper%20released%20by,these%20plans%20are%20on%20track. [4]http://www.legalserviceindia.com/legal/article-2867-artificial-intelligence-and-ipr.html [5]https://www.wipo.int/about-ip/en/artificial_intelligence/ip_administration.html [6]https://www.infoworld.com/article/3394399/machine-learning-algorithms-explained.html#:~:text=Machine%20learning%20algorithms%20are%20the,data%20set%20into%20a%20model.&text=depends%20on%20the%20kind%20of,the%20nature%20of%20the%20data. [7]https://searchenterpriseai.techtarget.com/definition/algorithmic-transparency [8]https://cyber.harvard.edu/events/2018/luncheon/01/monkeyselfie [9]https://www.wired.com/story/how-algorithm-favored-whites-over-blacks-health-care/ [10]https://www.reuters.com/article/us-amazon-com-jobs-automation-insight-idUSKCN1MK08G [11]https://blog.ipleaders.in/impact-artificial-intelligence-legal-system/ [12]https://www.weforum.org/agenda/2020/11/heres-how-technology-has-changed-and-changed-us-over-the-past-20-years/


ABOUT THE AUTHOR

This blog has been authored by Sneha Murthy, who is a 5th Year B.A., LL.B. (Hons.) student at MS Ramaiah College of Law, Bangalore.


[PUBLICATION NO. TLG_BLOG_21_10]