“By far, the greatest danger of Artificial Intelligence is
that people conclude too early that they understand it.”
— Eliezer Yudkowsky, American AI researcher.
What is artificial intelligence (AI)?
Since past few years, the idea of AI has been the talking point for the world. Contrary to popular belief, Artificial Intelligence (“AI”) is not a new concept, especially to the readers of science fiction (Sci-fi). In recent times, however, it is becoming more science and less fiction. The world of technology is changing rapidly, with computers and now robots, replacing simple human activities.
The term artificial intelligence is a combination of two words in a literal sense. The first, ‘Artificial’ means something which is not natural but man made. The second word “Intelligence” is quite difficult to define in a dictionary sense. However, Intelligence can be defined as an inclusive term which is relative to human mental capabilities i.e. perception, learning, planning, reasoning, problem solving, decision making, using language and learn from past experiences etc.
The Britannica Dictionary defines artificial intelligence as “the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings.”
How is AI different than the computers we had over the years?
Digital computers were first developed in the late 1940s. Created back then in the 20th century as a means to calculate, computers have come a long way from being just a number crunching machine; be it size — where multi-storied tower-sized computers have evolved into tiny pocket devices we call smartphones or be it speed —where computers of today are billions of times faster than the first computer ever made. Computers have affected our lives in such a profound manner that it’s hard to imagine the modern world without them. However, the rapid evolution of computers won’t stop at speed improvement and size reduction.
The key difference between the previous generation computers and the AI-based machines is the ability to think on their own and make decisions. The earlier computers were prone to follow what they were specifically programmed to do by the set of instructions fed into their memory system. However, the AI-based computers, as they evolve will be capable of decision making based on two primary variables. The first being result reached by calculative analysis using its resources both online and offline. Secondly, it will also learn from its past experiences using the machine-learning techniques. In simpler terms, it will be like how humans mold and adopt themselves with past experiences when faced with a choice of making decisions.
How much time will it take for AI to become main-stream?
It already has become prevalent in a number of activities. For instance, search engines such as Google, several medical diagnoses, voice and handwriting recognition tools are based on AI-based algorithms. There are chat-bots like haptik and virtual assistants like Google Assistant, Apple’s Siri, Samsung‘s Bixby & Amazon’s Alexa already existing and getting better every second. Google’s implementation of AI with respect to cross language translation is extremely accurate and to the point. Something which takes a whole lifetime for a human being to learn languages like Japanese, English, and Chinese and then translate those languages is being done in a matter of seconds by Google’s algorithms with remarkable accuracy. Similarly, AI is how Netflix is able to suggest a video that you might enjoy, Amazon is able to recommend a product that you may need next week based on past purchases, and it is also how your email filters spam messages that you do not want to receive in your inbox.
Further, there are various segments in the industrial space such as manufacturing, logistics, payment services, sales and customer relationship management (CRM) based applications which are adopting AI-based approach. In future, we may see automated transportation using driverless cars or even deployment of AI robots in hazardous and dangerous jobs or even a revolution in the sectors such as healthcare, power generation, education etc.
So what is the Legal Status of AI?
It can be said that the advancement of AI in India is at the stage of infancy. In the Indian context of law, the term AI has not been defined yet by any legislation or a judicial pronouncement. However, it can be said that, under the Indian laws, the status of a ‘person’ is not granted exclusively to human beings. The law has extended this status to non-human entities as well, whether they are corporations and other artificial legal persons.
As discussed above, no law currently in force in India recognizes artificially intelligent entities to be legal persons, which raises the question of whether there is a need for such recognition. While considering this question, one has to understand the essence of law when it considers a person separate legal entity. The basis of such recognition boils down to the basic three questions.
Firstly, whether such entity has the right to own a property?
Secondly, can this entity sue and be sued in its own name?
And finally, if the answers to the above two questions are affirmative then, can we make this entity subject to legal rights and duties?
On October 27, 2017, Saudi Arabia became the first country in the world which granted citizenship to a robot named ‘Sophia’. Sophia is a humanoid robot based on a neuron engine powered by AI. Sophia has the ability to follow faces, sustain eye contact and recognize individuals. She is also able to process speech and have conversations using a natural language subsystem. Its creator Mr. David Hanson has said that Sophia would ultimately be a good fit to serve in healthcare, customer service, therapy, and education.
Coming back to the question i.e. if AI is recognized as a person under the Indian Law, then how it will it fall within the existing ambit of laws such as intellectual property laws (Copyrights, Patents, Industrial Designs etc.), civil and criminal laws and taxation laws. For instance if an original literary work is produced by an AI-bot then will the copyright law provide protection to it? For the record, the section 2(d)(vi) of the Copyright Act, 1957 provides that author of the work “in relation to any literary, dramatic, musical or artistic work which is computer-generated, the person who causes the work to be created; In this case if the AI is recognized as a person then who shall be recognized as the owner of that copyright? . It could be argued that either the person who invented that AI-bot may have a claim on such copyright based on the definition provided under the act or since the AI-bot itself is a person then it shall be capable to hold the copyright of the same in its own name.
Similarly in case of a criminal offence, Will the AI-bot be liable for serving imprisonment Vis-à-vis like a human offender. These are the questions we may have to answer in near future. However, one thing is certain, If AI is allocated the status of a person then it will open up a Pandora’s Box as far as legal ramifications are concerned with respect to the laws in the country.
How will AI impact the financial markets?
AI has the potential of making a sea change in the entire financial market. These changes will affect the whole spectrum of the market including that of stock exchanges, market intermediaries, the market regulator, depositories, and investors. Following are the key areas where AI can make a groundbreaking difference.
- AI-based supercomputers can manage highly complex systems in an effective and autonomous manner. After providing the initial instruction sets, the AI-based system shall take over the handling of complex processes such Initial Public Offer (“IPO”), book building process, allotment procedure, stock settlement cycle etc. It shall also bring down the time required to complete the aforementioned procedures, meaning thereby quicker listing of securities and transfers.
- As we are aware, top-tier listed companies need to comply with several statutory compliances, AI-based tools will make it easier for the management to know the current status and deadlines of the compliances. Some companies are also in the process of developing of an AI-based internal accounting system will monitor, prevent and correct the flaws flowing from the books of the company, which in turn will prevent reoccurrence of frauds of high magnitude.
- Traders, Fund managers and analysts typically rely on fundamental analysis of securities. These methods are easy to learn and apply but not always effective. However, with AI-based tools, the analysis will combine the fundamental analysis with machine-learning based tools resulting in better portfolio suggestions to clients.
- Further, the analysis and research reports prepared by human beings are rather subjective and capable of bias. However, the analysis by an AI supercomputer will be more streamlined, consistent and reliable.
- AI-based computers can not only process a gigantic amount of data but also run analysis in real-time based on the results acquired, something which is not comprehensible for a human being in a limited time.
- Extensive use of AI will result in more predictable and efficient markets with lower volatility for extended periods of time with only occasional volatility spikes due to regime changes.
- From the viewpoint of the regulator, AI-based vigil mechanism systems will be much more effective than current process of combining humans with traditional computers at detecting the instances of fraudulent transactions such a securities fraud, IPO scams, front running, insider trading and other offences. AI-based tools will bring immense assistance during litigation to litigants as well as adjudicating tribunals and courts during the pendency of cases. An AI-based system can be developed which will understand the facts of the case and deliver the relevant applicable laws, judicial precedents and other research-based analysis autonomously. With the help of internet and having access to a nearly limitless amount of data, it will become highly difficult to suppress a material fact of the case.
- AI- based tools will be highly conductive for law firms as well, wherein tasks of due diligence, contract analysis, Technology-assisted review(TAR) for finding a particular fact in a document will become easier and enable lawyers work faster on a case. All this will allow lawyers to work into the creative, intellectual analysis that they are trained to do, for instance formulating litigation strategy and courtroom tactics.
- AI-based tools can review and suggest improvements in existing regulatory framework to the regulator itself. Such improvements shall be suggested by the AI on extensive and continuous analysis of the market conditions with respect both the internal and external factors affecting such market and the effectiveness of current regulatory framework governing the segment where change is suggested.
- Predictive actions are possible through use of technology especially Artificial Intelligence (AI) in utilizing the huge data available with the stock exchanges, regulators and courts by data management techniques. AI and Big Data Analytics through machine-learning can create patterns out of unstructured data such as the voluminous data sets. Say for instance in case of judiciary or regulatory perspective, Such large database can be used to predict the time, different types of cases take to be heard; what kind of outcomes result; how many cases are waiting to be heard.A modern system can be recording and making available such transparent and open data.AI can even understand contextual differences in meaning and help in predictive results. As an aid in traditional justice delivery mechanism, AI can thus play a supplementary role in the future. As a first step, it is important to realize the importance of databases, both structured and unstructured, and open up the same to analytics and technology.
- Some AI-based robo-advisers have been established by technology start-ups, such as Betterment and WealthFront in the U.S., Nutmeg in the U.K. and 8 Securities in Asia. While these independently operated robo-advisers account for only a small fraction of the overall global asset management industry, their presence has not gone unnoticed by incumbents. Such robo-advisers utilize algorithms to construct, manage and rebalance investment portfolios, typically using a pre-determined mix of exchange-traded funds (ETFs) to build low-cost, diversified, and liquid portfolios tailored to the investor’s objectives. The only thing the portfolio investor needs to do is to complete a profile that helps these robo-advisors to determine their risk appetite and/or investment objective, which is then used to build a portfolio for the investor.
- Finally, AI could support planning, coordinating, integrating, synchronizing, and directing activities to operate and defend stock exchange, regulator and intermediary networks and systems effectively, provide assistance in support of secure operation of such networks and systems, and enable action in accordance with all applicable laws, regulations and rules thereof.
So, what are the global and domestic initiatives taken by courts, government and regulators with respect to AI?
- The advent of AI is real. Even the judiciary is acknowledging its impact. Hon’ble Justice A.K. Sikri of the Supreme Court recently quoted “When we talk of Artificial Intelligence, it is in various fields where it is going to help like research. It helps not only in the research but in finding what the law is but right from the beginning where the problem comes, how it is to be solved, how it is to be tackled, and how the digital data can help us.” His colleague and another notable judge of the Hon’ble Supreme Court, Justice D.Y. Chandrachud remarked last year that “the idea of Artificial Intelligence is not to supplant the human brain or the presence of the judges but to provide a facilitative tool for the judges to reassess the processes which they follow or to reassess the work which they do with a view to ensure that their outcomes are more predictable, consistent and ultimately provide wider access to justice to the common citizens.”
- In the recent case of Matrimony.com Limited and Ors. vs. Google LLC and Ors., Competition Commission of India (CCI) observed that data’s importance in this centuryis going to be on par with what oil was to the last one. The commission noted that using the AI-based innovation; such data can be used extensively for revenue generation.
- During the budget 2018 speech, Arun Jaitley, the union finance minister announced a national program on AI to be spearheaded by Niti Aayog. A high-level committee headed by Niti Aayog vice-chairman Rajiv Kumar has been formed to lay out a roadmap for India’s R &D on AI.
- Recently Minister of State for Corporate Affairs Shri P P Chaudhary claimed that AI-based technology was being considered for integration within the MCA21 portal. The primary purpose of the technology will be to detect finer discrepancies in details provided by the companies in the balance-sheets. Such artificial intelligence software then would match the data submitted to the ROC over the time with the balance-sheet of the company. If there is any discrepancy, action can be taken against the companies and their directors in the interest of small investors. Further, such data can also be used to detect shell companies which have infested the listed space.
- In September 2017, a Policy group" was created by the Ministry of Electronics and Information Technology with representation from the academia, which has done a lot of research on the subject, and Nasscom for the industry's perspective. The group will make its observation and recommendations on aspects like skilling the workforce, privacy, security and fixing responsibility if anything goes wrong with the usage of AI.
- Japan has recently announced that they will start the usage of AI-based technology to manage the government pension scheme with effect from fiscal year 2018, initially on a trial basis. The reason for such action is deemed to be quicker and more comprehensive analysis by AI than human counterparts while being cost effective at the same time.
- China has laid down plan to become the world leader in AI by 2030. The state council of the communist republic has released a policy of intent detailing capital-intensive investments by companies, military and government in AI technology.
- The United States of America (USA) has been one of the early adopters of AI phenomena. In 2016, The White House released a report detailing the future usage of AI. America is already applying the AI mechanism to to predict medical complications, reduced traffic wait times and even to analyze and comprehend tourist data.
- In September 2015, in response to some Canadian registered portfolio managers and restricted portfolio managers starting to operate as “online robo-advisers”, the Canadian Securities Administrators (CSA) clarified in CSA Staff Notice 31-342 that there is no ”online advice” exemption from the normal conditions of registration for a portfolio manager. The rationale behind the clarification is that Since the law is technology neutral, and the obligations applying to the provision of traditional (i.e. non-digital) financial product advice and digital advice are the same. Robo –advisers also fall within the purview of the existing legal framework.
- NSE, the country’s largest stock exchange has stressed the need for technological innovation including the usage of AI. In an interview, Mr Vikram Limaye, CEO of the bourse noted that AI can be utilized to assist human employees in making better-informed decisions, acceleration of low-risk tasks and to facilitate more transparent algo and high frequency trading (HFT)
But, what about the warnings against AI?
Yes, it is quite true that some eminent scientists and entrepreneurs have warned against the technological advancement of AI and their stance is probably valid in case of usage of AI in warfare. However, opposing AI even in sectors wherein it is primarily used for public good is far-fetched. Since AI is in its beginning right now, extensive testing of its working is required. Such testing can be achieved using trial and error method with the aid of human supervision. As AI learns from its past experiences including mistakes, over the time, it will become less and lesser prone to make mistakes, thereby may be becoming flawless at one point. Further, AI is not subject to typical human bias and emotions which will result in consistent and objective approach in dealing with matters. However, it needs to be understood that AI needs good data. If the data is incomplete or biased, AI can exacerbate problems of bias and unrealistic results.
Another instance why AI is opposed is due to threat that it will advance its use at the cost of jobs of human beings. While it is true that AI will subsume some of the tasks human beings are doing, but most of the jobs which AI might kill will be on the lower side of the job chain.
AI divides opinions, some are willing to adopt it, and some are vehemently opposing it. Another threat AI faces is the possibility of being hacked and misused for unlawful purposes.
Primarily, there are two competing theories which predict the AI’s future but there is no certainty as to how AI will evolve.
- AI can never reach a point where it is superior to humans
- AI will surpass the human participation in every single field.
AI’s application is developing at an increasingly faster pace, creating new opportunities to achieve better outcomes for various stakeholders but at the same time, as with any change, new risks and vulnerabilities may arise. As far as the disadvantages are concerned, several steps to safeguard the AI mechanism are suggested. One check which is suggested that is to keep human in the loop of the AI process by way of external verification whenever we can, so as to check and correct if the algorithms’ guesses are incorrect. Secondly, it is suggested to build robust cyber security mechanisms. One interesting argument presented by security experts is to program AI specific algorithms to safeguard itself from the activity of hackers and since AI can think on its own, it is suggested that it will evolve its instruction sets on its own to counter the newer threats. As far as jobs are concerned, it’s similar to how many jobs were terminated when computerization happened back at end of 20th century. But does it mean we shouldn’t have moved on from papers to computers back then? However, some analysts suggest that since AI is evolving consistently, it will soon replace every single job. Such an act cannot be ruled out but only time will tell if that does happen. Further, the development of AI bots who rival the cognitive intelligence of human beings is still decades away according to scientists For instance, The Robo-Advisers are still at a nascent stage wherein there are several questions on their usage due to the possibility of error in the calculations, over simplistic generalized approach and lack of personal touch which an investor gets from a traditional human being.
What is the solution?
Education transformation is one area which needs urgent intervention. The future generation needs to prepare for a future wherein they will not only compete with humans but highly intelligent machines as well. It is time to accept that automation has come here to stay and is bound to create a future for all of us. The need-of- the-hour is to design governance mechanisms that ensure AI and its allied technologies do not transgress boundaries. The Government has several roles to play. It needs to convene conversations about important issues and help to set the agenda for public debate. It should monitor the safety and fairness of applications as they develop, and adapt regulatory frameworks to encourage innovation while protecting the public. After all, it is true that the technology is only as good as its creator. In a nutshell, it can be concluded that AI is going to be a disruptive development in technology. The ramifications will only be visible once the time takes it course. After all, we have barely scratched the tip of this iceberg.
— by Hritik Bhimani, Advocate....