India’s Artificial Intelligence Roadmap

[This piece was originally published on the Centre for Communication Governance blog on June 25, 2018]

There is now a near universal perception that Artificial Intelligence technologies are set to disrupt every sphere of life. However, this is coupled with concern regarding the social, ethical (and even existential) challenges that AI might present. As a consequence, there has been an uptake in interest by governments on how best to marshal the development of these technologies. The United Kingdom, the United States, China, and France, among others, have all released vision documents that explore these themes.

This post presents a brief overview of such initiatives by the Indian government.

Task Force on Artificial Intelligence

In August 2017, the Ministry of Commerce and Industry set up a ‘Task Force on Artificial Intelligence for India’s Economic Transformation’. A panel of 18 members was formed with the objective of exploring how Artificial Intelligence could be best deployed in India.

The Task Force released its Report in May 2018, where it characterized AI as a ‘socio-economic problem solver at a large scale’, rather than simply a booster for economic growth. It sought to explore domains which would benefit from government intervention, with the objective of improving quality of life, and generating employment. The report identifies 10 sectors where AI could be deployed — Manufacturing, FinTech, Healthcare, Agriculture and Food Processing, Retail, Accessibility Technology, Environment, National Security and Public Utility Services. It attempts to identify challenges specific to each sector, as well as enabling factors that could promote the adoption of AI.

The report also explores the predicted impact of AI on employment, as well as other broader social and ethical implications of the technology. It concludes with a set of recommendations for the government of India. A primary recommendation is to constitute an Inter-Ministerial National Artificial Intelligence Mission (N-AIM) with a 5 year budget of Rs. 1200 Crores. Other recommendations focus on creating an ecosystem for better availability of data for AI applications; skilling and education initiatives focused on AI; standard setting, as well as international participation in standard setting processes.

NITI Aayog’s National Strategy for Artificial Intelligence

In his Budget speech, the Finance Minister had tasked the NITI Aayog with formulating a national programme for Artificial Intelligence. In June 2018, the NITI Aayog released its roadmap in the form of the National Strategy for Artificial in India.

The paper frames India’s AI ambitions in terms of increasing economic growth, social development, and as an incubator for technology that can cater to other emerging economies. It focuses on 5 sectors as avenues for AI led intervention. These are healthcare, agriculture, education, smart cities, and smart mobility. It also identifies some key challenges to the effective adoption of AI. These include low awareness, research, and expertise in AI along with an absence of collaboration; the lack of ecosystems that enable access to usable data; high resource costs; and ill-adapted regulations.

The paper then presents a series of recommendations to address some of these issues. In order to expand AI research in India, it proposes a two-tier framework to focus on basic research as well as application based research. It also proposes the creation of a common computing platform in order to pool cloud infrastructure, and reduce infrastructural requirements for such institutions. It further suggests a review of the intellectual property framework to enable greater AI innovation. In order to foster international collaboration, the paper proposes the creation of a supranational CERN-like entity for AI. It also recommends skilling and education initiatives to address job creation, as well as the current lack of AI expertise. In order to accelerate adoption, it proposes a platform for sharing government datasets, along with a marketplace model for data collection and aggregation, for data annotation, as well as for deployable AI models.

The paper concludes with its recommendations for ‘responsible’ AI development. It recommends that there be a consortium of the Ethics Councils at each of the AI research institutions. It further proposes the creation of a Centre for Studies on Technology Sustainability. It also emphasizes the importance of fostering research on privacy preserving technology, along with general and sectoral privacy regulations.

Further reports suggest that a task force will be set up to execute the proposals that have been made, in coordination with the relevant ministries.

MeitY Committees

It has also been reported that four committees have been constituted in February 2018 to deliberate on issues of ‘data for AI, applications of AI, skilling and cyber security/legal, ethical issues.’ However, there have been no reports about when the committees will present their recommendations, and whether they will be made available to the public.

~

India appears to be at the nascent stage of formulating its approach towards Artificial Intelligence. Even so, it is encouraging that the government recognizes the importance of its stewardship. Purely market led development of AI could imply all of its disruption, without any of the envisaged social benefits.