Thus, the adoption of artificial intelligence is not the challenge but its ‘regulation’ is a slippery slope. Which leads us to questions such as whether we need to regulate artificial intelligence at all? If yes, do we need a separate regulatory framework or are the existing laws enough to regulate artificial intelligence technology?

Artificial intelligence goes beyond normal computer programs and technological functions by incorporating the intrinsic human ability to apply knowledge and skills and learning as well as improving with time. This makes them human-like. Since humans have rights and obligations, shouldn’t human-likes have them too?

But at this point in time, there have been no regulations or adjudications by the Courts acknowledging the legal status of artificial intelligence. Defining the legal status of AI machines would be the first cogent step in the framing of laws governing artificial intelligence and might even help with the application of existing laws.

A pertinent step in the direction of having a structured framework was taken by the Ministry of Industry and commerce when they set up an 18 member task force in 2017 to highlight and address the concerns and challenges in the adoption of artificial intelligence and facilitate the growth of such technology in India. The Task Force came up with a report in March 2018[1] in which they provided recommendations for the steps to be taken in the formulation of a policy.

The Report identified ten sectors which have the greatest potential to benefit from the adoption of artificial intelligence and also cater to the development of artificial intelligence-based technologies. The report also highlighted the major challenges which the implementation of artificial intelligence might face when done on large scale, namely (i) Encouraging data collection, archiving and availability with adequate safeguards, possibly via data marketplaces/exchanges; (ii) Ensuring data security, protection, privacy and ethical via regulatory and technological frameworks; (iii) Digitization of systems and processes with IoT systems whilst providing adequate protection from cyber-attacks; and (iv) Deployment of autonomous products and mitigation of impact on employment and safety.[2]

The Task Force also suggested setting up of an “Inter–Ministerial National Artificial Intelligence Mission”, for a period of 5 years, with funding of around INR 1200 Crores, to act as a nodal agency to coordinate all AI-related activities in India.

Core Legal Issues

When we look at the adoption of artificial intelligence from a legal and regulatory point of view, the main issue we need to consider is, are the existing laws sufficient to address the legal issues which might arise or do we need a new set of laws to regulate the artificial intelligence technologies. Whilst certain aspects like intellectual property rights and use of data to develop artificial intelligence might be covered under the existing laws, there are some legal issues which might need a new set of regulation to overlook the artificial intelligence technology.

• Liability of Artificial Intelligence

The current legal regime does not have a framework where a robot or an artificial intelligence program might be held liable or accountable in case a third party suffers any damage due to any act or omission by the program. For instance, let us consider a situation where a self-driven car controlled via an artificial intelligence program gets into an accident. How will the liability be apportioned in such a scenario?

The more complex the artificial intelligence program, the harder it will be to apply simple rules of liability on them. The issue of apportionment of liability will also arise when the cause of harm cannot be traced back to any human element, or where any act or omission by the artificial intelligence technology which has caused damage could have been avoided by human intervention.

One more instance where the current legal regime may not be able to help is where the artificial intelligence enters into a contractual obligation after negotiating the terms and conditions of the contract and subsequently there is a breach of contract.

In the judicial pronouncement of United States v Athlone Indus Inc[3] it was held by the court that since robots and artificial intelligence programs are not natural or legal persons, they cannot be held liable even if any devastating damage may be caused. This traditional rule may need reconsideration with the adoption of highly intelligent technology.

The pertinent legal question here is what kind of rules, regulations and laws will govern these situations and who is to decide it, where the fact is that artificial intelligence entities are not considered to be subject of law.[4]

• Personhood of Artificial Intelligence Entities

From a legal point of view, personhood of an entity is an extremely important factor to assign rights and obligations. Personhood can either be natural or legal. Attribution of personhood is important from the point of view that it would help identify as to who would ultimately be bearing the consequences of an act or omission.

Artificial intelligence entities, to have any rights or obligations should be assigned personhood to avoid any legal loopholes. “Electronic personhood”[5] could be attributed to such entities in situations where they interact independently with third parties and take autonomous decisions.

• Protection of Privacy and Data

For the development of better artificial intelligence technologies, the free flow of data is crucial as it is the main fuel on which these technologies run. Thus, artificial intelligence technologies must be developed in such a way that they comply with the existing laws of privacy, confidentiality, anonymity and other data protection framework in place. There must be regulations which ensure that there is no misuse of personal data or security breach. There should be mechanisms that enable users to stop processing their personal data and to invoke the right to be forgotten. It further remains to be seen whether the current data protection/security obligations should be imposed on AI and other similar automated decision-making entities to preserve individual’s right to privacy which was declared as a fundamental right by the Hon’ble Supreme Court in KS Puttaswamy & Anr. v Union of India and Ors[6]. This also calls for an all-inclusive data privacy regime which would apply to both private and public sector and would govern the protection of data, including data used in developing artificial intelligence. Similarly, surveillance laws also would need a revisiting for circumstances which include the use of fingerprints or facial recognition through artificial intelligence and machine learning technologies.

At this point in time there are a lot of loose ends to be tied up like the rights and responsibilities of the person who controls the data for developing artificial intelligence or the rights of the data subjects whose data is being used to develop such technologies. The double-edged sword situation between development of artificial intelligence and the access of data for further additional purposes also needs to be deliberated upon.

Concluding Remarks

In this evolving world of technology with the capabilities of autonomous decision making, it is inevitable that the implementation of such technology will have legal implications. There is a need for a legal definition of artificial intelligence entities in judicial terms to ensure regulatory transparency. While addressing the legal issues, it is important that there is a balance between the protection of rights of individuals and the need to ensure consistent technological growth. Proper regulations would also ensure that broad ethical standards are adhered to. The established legal principles would not only help in the development of the sector but will also ensure that there are proper safeguards in place.