With the rapidly developing digital world, machines are becoming a vital element for interactions with humans, thus leading to innovations in artificial intelligence (AI). AI has become an integral part of almost all industries, making its way into healthcare, education, and transportation to food supply, energy, environmental management and many more. The potential of AI to become the backbone of the information age in the coming years presents us with few challenges, primary one being the ethical considerations around it. Understood as a set of values, principles, and techniques that employ widely accepted standards of right and wrong to guide moral conduct in the development and use of AI technologies, it seem to have occupied the centre-stage in discussions around AI today.
AI and its ethical violations have been a source of concern for most regulators around the world. Recently, United States, came under the regulatory radar inventing an algorithm which created racial bias. Termed as ‘Optum’, the algorithm is sold by a leading health service company and is alleged to have been advising doctors and nurses to give white patients more attention than black patients. On similar lines, there have been allegations about algorithm developed by Goldman Sachs, which tends to create gender bias by granting men more credit limits than women. Along with these cases, there are many more, which includes the infamous Cambridge-Analytica scandal, where Facebook was investigated for providing access to users’ personal data. The US Department of Housing and Urban Development recently sued Facebook, alleging that its ad-serving algorithms permitted discrimination based on gender and race.
Amidst the global debate on ethical considerations around AI, India too seems to have initiated measures to address this issue. While India lacks a comprehensive guidance framework for the usage of AI systems. However, certain industry-specific policies have been introduced to address this issue.
India’s Policy Response
There is currently no overarching AI legislation in India. The only legislation to protect data of users/customers from been unethically used by algorithmic platforms is the Information Technology Act, 2000 (IT Act), followed by the Personal Data Protection Bill, 2019 (PDP Bill).
The IT Act has served as the cornerstone of India's data protection laws. The IT Act's provisions, when combined with the Information Technology (Reasonable Security Practices and Procedures and Sensitive Personal Data or Information) Rules, 2011 (SPDI Rules), create a technology-neutral framework for the protection of sensitive personal data for all bodies corporate. The newly enacted PDP Bill intends to provide for a comprehensive legislation, outlining various aspects of privacy protections, with which AI solutions must comply. It addresses data processing limitations, security safeguards to protect against data breaches, and the inclusion of special provisions for vulnerable users such as children. Furthermore, the PDP Bill calls for a vibrant data protection legislation in which the law is supplemented with regulations and codes of practise, making it easier for privacy to evolve in tandem with evolving technologies.
Coupled with these legislations, several measures were taken to address and adapt to the fast-developing AI market in India. In January 2019, Securities and Exchange Board of India (SEBI) introduced a circular to market intermediaries to conform to certain reporting requirements for AI and Machine Learning applications. The reporting is aimed at compiling a list of AI systems and in guiding the future policy. In the healthcare sector, to assure trustworthiness of AI technologies, the National Digital Health Mission (NDHM) strategy recognises the need for guidelines.
In August 2017, the Union Ministry of Commerce and Industry established an Artificial Intelligence Task Force with the goal of ‘embedding AI in our Economic, Political, and Legal thought processes so that there is systemic capability to support the goal of India becoming one of the leading AI-rich economies.' The Task Force's work aims at providing light on the route that India's AI strategy should take. Its primary strength is its concentration on accessibility technologies. However, a cursory (at most) ethical and social examination of India's AI ecosystem reveals the lack of legal, policy, and civil society engagement in this process.
In February 2018, the Union Ministry of Electronics and Information Technology (MeitY) set up four groups to establish an AI national plan. The four committees are focused on AI in the context of citizen-centred services, data platforms, skills and research, and legal, legislative and regulatory prospects and cyber security.
In order to guide government efforts, the National Institution for the Transformation of India ('NITI Aayog'), a government-run think tank, has produced a National AI policy. NITI Aayog worked with Google to train and incubate start-ups to create and integrate AI-based solutions in their business models in early May 2018 to enhance economic productivity. NITI Aayog's Data Empowerment and Protection Architecture (DEPA) provides a technology framework for retention of control and leverage by people of their personal data for the use of services and benefits.
The Way Forward
While there have been efforts towards formulating a comprehensive AI policy for India, most of the developments have been disjointed. India's AI development has so far been a disjointed policy process. There is no denial that measures were taken to address the issue of AI in India. However, all existing efforts are largely aimed at promoting economic development and growth. While the NITI Aayog study takes into account societal requirements, it falls short of a full examination.
AI is generally regarded as a business potential. This is a critical topic for Indian AI policy because these systems have the ability to reshape economies. However, it is not the only factor that necessitates resources and in-depth knowledge. The ethical, social, and legal ramifications of AI on daily life must be considered on an equal footing, especially because their impact is sometimes irreversible, despite being impossible to show or prove.
Second, AI's ongoing research and deployment methods appear to be aimed at enhancing innovation, which can lead to a false dichotomy between ethics and innovation. Ethics may and should be entrenched as a key feature of breakthrough technologies, but this has yet to be generally acknowledged and accepted.
Third, present AI programmes largely seek participation from government and industry, with a little amount of participation from academia, but no participation from civil society. There was no open call for participation in these processes or committees, which were appointed or nominated unilaterally. These risks removing social and ethical considerations from the forefront of debates, whereas an AI approach that incorporates these factors from the start is significantly more viable. Given that the AI ecosystem is complex, interconnected, and relies on collaboration and partnership between multiple stakeholders, these risks restricting the effectiveness and reach of AI policy in India.
Fourth, the limitations of automated processing and technical solutions for societal and systemic problems are frequently overlooked. The ethical and social implications of data-driven decision-making, as well as its inherent hazards and limitations, must be redefined as a fundamental factor in India's AI policy development. The mere fact that a decision is based on evidence does not imply that it is fair, just, or suitable.
Fifth, while state use and acceptance of AI for law enforcement, defence, and national security is on the rise, policy debates have so far failed to address the parameters within which such usage should take place. Even when national security is mentioned, such as in the AI Task Force Report, the responsibilities and limits on state power are not discussed. More transparency and critical dialogue are needed in the interaction between individuals and the state, as well as the effects of AI on the exercise of rights.
It is important that the above issues are taken into consideration while formulating a policy for AI in India, especially when it has integrated itself into almost all sectors.
About the author
Shatabdi Nayak is pursuing B.A., LL.B at Damodaram Sanjivayya National Law University Visakhapatnam.