Ritika Sharma explores the legal ramifications of a world dominated by artificial intelligence
Technology is changing the world around us with increasing rapidity. New technologies are being incorporated more and more into our daily lives. Computer programs, algorithms and robots are replacing simple human activities. And artificial intelligence (AI) lies on the cutting edge of the technological spectrum in the industry today. Simply put, AI is the capability of a machine to imitate intelligent behaviour. The era of AI has begun, whether we like it or not.
Computer scientist John McCarthy coined the term “artificial intelligence” at a Dartmouth Conference at the Massachusetts Institute of Technology, defining AI as the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable. It’s an umbrella term that refers to information systems inspired by biological systems, and encompasses multiple technologies including machine learning, deep learning, computer vision, natural language processing, machine reasoning, and strong AI.
Is AI regulated?
AI is poised to have a transformative effect on numerous industries, and is already permeating through all aspects of society, and in our everyday lives. AI adoption is so widespread that hardly any sector or industry is untouched, from education, healthcare, transportation, defence and security, and virtual reality to decision making and policy making.
AI will probably play a major role in shaping the evolution of various industries. But, if it is not regulated properly, it could lead to unmanageable implications, and it’s therefore not surprising that the Government of India is looking to regulate AI with a proper legislative framework to ensure that AI tech is properly developed and implemented.
The absence of a regulated framework for AI is not an option. For instance, let us consider a hypothetical scenario: If a driverless car running on an AI algorithm causes a human fatality or damage to public or private infrastructure, then who is to be held accountable for the damages or consequences? Given that AI is not a legal person, will there be no action against anybody? Given that Saudi Arabia has accepted an AI-based robot, Sophia, as its citizen, can other countries follow suit and indict AI for such crimes? These situations raise more questions than answers.
Countries around the world are discussing how to regulate AI along with their increasing awareness of the potential economic and social benefits of deploying AI and AI-based applications. Numerous states have taken progressive strides in terms of preparing policies in the past couple of years, and China, the UK, France and Japan have all published their respective strategies recently.
The Constitution of India devises the legal framework allocating rights and obligations to persons and citizens of India, but the courts of India are yet to adjudicate upon and determine the legal status of AI. India can learn from the experiences of countries such as Japan, the US and China, which are all in more advanced stages of regulation due to the higher penetration of these technologies.
In August 2017, India’s Ministry of Industry and Commerce has constituted the 18-member Taskforce on AI for India’s Economic Transformation – consisting of experts and government bodies including the National Institution for Transforming India’s (NITI) policy commission, the Ministry of Electronics and Information Technology, the Department of Science and Technology, the Unique Identification Authority of India (UIDAI), and the Defence Research and Development Organization – to ascertain the possibilities of making use of AI-based applications for development across various sectors. This taskforce recently released a report detailing certain recommendations that they made to the Ministry of Industry and Commerce on formulation of an AI policy in India.
Despite a lot of deliberation on the technological aspects, no action has yet been taken to address the immediate hurdle; a legislative framework governing AI is a must if it is to be beneficial to society.
A legislative framework
Due to the transformative nature of AI-based technologies, they have the potential to pose multifarious legal challenges in the future. It is challenging to predict how the legal sector will function with advances in AI and vice versa depending on social, cultural, economic and other factors. Even though the requirements for formulating and developing a legislative framework concerning AI may be numerous, there are prominent legal issues to address.
First and foremost, there is no existing law that recognizes AI as a “legal person”. Given that Sophia has been granted citizenship in Saudi Arabia, it becomes extremely important that we establish a jurisprudence around “legal personality” in India as well, such that the specific rights and obligations of an AI can be demarcated.
Another issue emanating from this one is that as long as the legal personality of an AI is not established, it is impossible to attribute any form of liability to an AI, in case of any harm inflicted, or a crime committed, by an AI. Until the liability of an AI or the legal personhood of an AI is pronounced, any contract involving an AI presence can be questioned, as the same would not be regarded as a valid contract in India.
There are also significant issues concerning the volumes of data stored by an AI. This issue gains particular significance in light of the right to privacy, which was recently granted the status of a fundamental right under the Constitution of India. It is therefore needless to emphasize the importance of formulating a set of rules or guidelines to regulate data stored and used by AI. The Personal Data Protection Bill, 2018, is already in place and perhaps the same could be augmented to accommodate this issue.
Impact on legal sector
The history of humanity is testament to technology reshaping the way we do things, and AI is proving no different. Until recently, lawyers used to rely on the ability of an adept paralegal for sifting through data and pointing out irregularities or wording to work upon while drafting major agreements. But, as the complexity of laws and the rate of data capture increases, the scope and human ability to sift through these mountains of data have been increasingly put into question. Today, the legal domain faces a glaring bottleneck, namely the lack of the ability to process large quantities of data sets on a daily basis.
The stakes are higher than ever, with little or no margin for human error. Compound that with the ever-changing legal landscape, and local regulations and judgments, and it is increasingly difficult for human-based interventions to update legal paraphernalia across documents and transactions. In other words, in the legal fraternity as with others, the data explosion is proving too much for humans to handle.
AI seeks to answer such problems for the modern enterprise. AI can help the legal sector by providing major insights within a matter of seconds. For instance, by automatically updating the documents or providing the latest regulatory rulings, these algorithms are increasingly making human-based interventions obsolete. Not only does AI have the advantage of being error-proof, it is proving to be cheaper compared to the costly man-hour rates charged by leading law firms across the globe. There are certain AI software programs that have been developed to harness the newest enabling technology, in combination with domain expertise, to provide legal professionals with greater clarity, efficiency and understanding of agreements that they confront most frequently in practice. Furthermore, there are many uses of such software and technology; the forms and clauses are assembled from the automated analysis of thousands of negotiated agreements, providing lawyers with an immediate sense of which clauses tend to be heavily negotiated and which are not.
What is unsettling, however, is that the legal fraternity is merely part of a larger trend of cross-industrial automation powered by AI-based technologies. With manufacturing, pharmaceutical and other industries increasingly adopting this new global standard of data synthesis, the legal fraternity cannot afford to fall behind. The adoption of AI in the legal fraternity is a natural stepping stone along its evolutionary path.
To sum up, AI provides excellent opportunities in the legal fraternity including, but not limited to, reviewing agreements, conducting legal research and even predicting the probable outcome of certain cases being adjudicated before the courts. AI-related tools also aid in case-law searches, or search and registration of an intellectual property right. There are some legal offices that are also making use of AI-based applications to help prepare invoices and arrive at an accurate billing, which is an arduous task for law firms.
In these circumstances, the use of AI-based tools may prove to be an innovation in the way the legal sector operates in the times to come. However, there remain many nuances that are still uncertain and difficult to predict.
The dependency of humans on machines is only prone to increase with time, so it is important to ensure that current concerns with regard to AI are adequately addressed. It is highly recommended that draftsmen and regulators are cognizant of the fact that a perfect balance has to be maintained between protecting the rights of the people, and fostering advancement of technology in the country. Failure to strike this balance may have an adverse effect on either of the two; the spirit of using such technology should be to ensure that AI proves to be a boon, not a bane.
Ritika Sharma is manager (M&A) – legal at Cipla.