What India Needs to Do to Build a Robust AI Ecosystem in the Country

There are two broad requirements for diffusion: policy push for spread of AI adoption, and trust of consumers or users, which is the 'policy pull'.

In December 2018, the Government of India approved the Rs 3,660 crore national mission on cyber-physical system technologies: artificial intelligence (AI), internet of things (IoT), machine learning (ML), deep learning, big data analytics, robotics, quantum computing, quantum communication, quantum encryption, data science and predictive analytics.

Earlier the same year, NITI Aayog had published the National Strategy for Artificial Intelligence discourse paper, which recommended setting up an AI Research, Analytics and knowledge Assimilation (AIRAWAT) platform. Meanwhile, the government had also constituted four committees on AI in four well-considered areas. These committees have since submitted their reports to the government; they are available on the MeiTY website for public comments.

India is gradually adopting artificial intelligence into its system. Indian startups and unicorns have been deploying AI tools for better delivery of services. Many of the recent initiatives of the government, whether the Aarogya Setu app or the MyGov application, aim use AI to make world class products. However, we are still miles away from unlocking the true value of the huge repository of data collected from different sources using AI and ML techniques, in both the government and the private sector.

Improving diffusion of AI

There are two broad requirements for diffusion: policy push for spread of AI adoption, and trust of consumers or users, which is the ‘policy pull’. The policy push has been inadequate. The other key ingredient of diffusion is to gain the trust of consumers and citizens by protecting their privacy and data, which is again missing due to the absence of a privacy law in India.

First, there has to be increased long-term investment in AI in both the public and private sectors. India lags behind the top five geographies for private sector investment in AI. The US is far ahead, with investments worth $18 billion, followed by Europe ($2.6 billion) and Israel ($1.8 billion). According to one estimate, China accounted for a greater share of global private-sector funding than the US.

Large companies can afford to invest in R&D but only a few are venturing into it; they are either scared, being risk averse, or are slow. The start-ups don’t have the money to invest in developing AI and are facing capacity issues. Accordingly, they prefer using existing models rather than develop new ones.

Second: the scope of using AI and blockchain to unlock efficiencies of organisations and systems is maximum in our public sector and government departments. To boost AI infrastructure in India, there is an urgent need for government departments both at the Central and the state levels to adopt AI-based solutions to solve the legacy and intractable problems being faced by them – from increasing agricultural productivity to climate change, from air pollution to disaster management. This will have a multiplier effect that could spur the private sector to jump in.

Third: the government and the private sector must work hand-in-hand, particularly on the R&D aspects of AI. Public-private partnership models bringing together the expertise of research institutions and universities (including the institutions of eminence), market knowledge along with financial skin-in-the-game by the corporations and partial financial grants by the government will inspire confidence to innovate using AI. The involvement of the private sector could ensure the research output is market-competitive.

Fourth: promoting STEM courses in early education, revisiting the course-curriculum in colleges and on-boarding industry experts to teach final year students were some of the key recommendations by a tech policy think-tank, The Dialogue, sometime back. Indian higher education programmes also need to be tweaked to include training on AI that cuts across disciplines. The vast network of hugely successful and talented Indian-origin professionals heading large multi-national corporations abroad must be utilised to institutionalise knowledge-sharing, mentoring and collaborative R&D initiatives.

Fifth: a fundamental challenge with regards to the development of AI in India is the quality of connectivity in the country. India needs to work towards improving its infrastructure and thereby allow for an environment that supports the growth of new technology.

Last: because AI is so powerful as it has all the information about a consumer to potentially be able to manipulate their behaviour, choices and preferences, India immediately requires a data protection law to ensure individual privacy.

AI is the future that India can ill-afford to ignore to solve some of the most intractable problems of its people. Fortunately, it has potential in its labour force, academic institutions and geopolitical status. If recognised and used correctly, India can become a leader in AI development. Further, indigenous innovation can be incentivised by creating a ‘Central Innovation Fund’ accessible to both the central and the state Governments to propagate AI in different sectors.

In order to guide India’s AI missions, we may also like to create a body like Indian Space Research Organisation, which has successfully guided India’s space missions with a high degree of self-reliance. Thus far, the sectoral approach has led to an arguably silo-type approach towards AI, which must be replaced with a whole-government approach like that of Singapore to foster a holistic and trustworthy AI ecosystem in India.

Amar Patnaik is Member of Parliament, Rajya Sabha, from Odisha, a former CAG bureaucrat with a master’s degree in public management from the Lee Kuan Yew School of Public Policy, Singapore, and the Kennedy School of Government, Harvard University