we seek to implement new AI-based technologies
to improve coordination and decrease transaction
cost. However, this may not be financially optimal
or efficient for the private sector, so a sustained government intervention is needed to improve market
conditions for farmers and other actors.
We will also have the governance stack. There are
many different welfare schemes (about 434) in India,
with about $60 billion worth of payouts, where the
amounts are dispersed directly to the bank accounts
through direct benefit transfer.
Editorial Team: How will the government protect
Arvind Gupta: India is coming up with a formal
your data and you give consent to people for your
data. Now we are formalizing and providing a legislative framework for supporting this. India is following a very different approach than GDPR. We
are trying to balance data empowerment, access,
anonymity, availability, privacy, and innovation,
which we call the data empowerment and protection architecture. 10
Editorial Team: What is India doing to build and
sustain a workforce that is proficient in building and
adopting AI technologies?
Arvind Gupta: We need more core science, technology, engineering, and mathematics research in
colleges. Also, what we really need is convergence. AI
in isolation will not solve problems. We need to develop medical facilities integrated with tech and research. We need to focus on interdisciplinary research
that involves areas like health and agriculture. They
all need to come together. To solve this problem, we
are partnering with AI companies to build capacity
at select tech campuses such as IITs and liberal arts
colleges. India has excellent research facilities and a
large workforce that is trained in IT. This talent pool
can be considered the greatest strength of our country
in the development of an AI industry. We need to
create schemes under which the two come together,
so the cutting-edge research that is done for example
in our IITs can be leveraged to create real world applications and global intellectual property based on
AI. Access to this research and collaboration between
scientists and entrepreneurs is an important step to
facilitate innovation and growth in this field.
Interdisciplinary research is key to ensure the thorough understanding of problems and developing the
right technology and framework to address them.
Editorial Team: What types of regulatory issues are
anticipated from the use of AI technology in India?
Arvind Gupta: Issues of liability, intellectual property,
and accountability will be major concerns in helping
AI become mainstream. These are very serious issues,
and I think harm has to be defined first of all, and also
liability, and best-effort basis. A lot more thinking is
required on questions such as: Where is the ownership?
Who is liable for what? Where does one’s loyalty lie?
We have already seen these issues come up with
autonomous vehicles. Another example is in using
AI for medical diagnostics. Should the government
take action if a patient is harmed by the usage of
an AI model? Or should the government ensure that
such scenarios do not occur? We need to be able to
determine who is liable in situations where there are
a lot of codependencies between actors.
An international consortium must be created to set
standards for safeguards that are to be respected in
the development of AI applications. AI use-cases that
have the potential to discriminate against or harm in-
dividuals must be subject to prior testing and, where
the safety of citizens cannot be guaranteed, the gov-
ernment must step in.
Editorial Team: Thank you, Arvind, for taking the
time to share your perspectives with us and the AI
Arvind Gupta: You’re welcome. It was my pleasure.
2. See timesofindia.indiatimes.com/trend-tracking/digital-india-to-push-ai-ml-blockchain-doubles-allocation-to-rs-
3. See yourstory.com/2018/07/artificial-intelligence-can-
4. See economictimes.indiatimes.com/industry/cons-products/
5. See niti.gov.in/writereaddata/files/document_publication/
6. The Ayushmaan Bharat is available at www.india.gov.in/
7. Information about the National Health Stack is available
9. See timesofindia.indiatimes.com/business/india-business/
10. Information about the data empowerment and protection architecture is available at indiastack.org/depa/.
Arvind Gupta is chief executive officer of MyGov, Government of India.
Biplav Srivastava is a distinguished data scientist and master
inventor at IBM’s Chief Analytics Office.
Daby Sow is a principal research staff member and research
manager at IBM T.J. Watson Research Center in Yorktown
Heights, New York.
Ching-Hua Chen is a research staff member at the IBM T.J.
Watson Research Center in Yorktown Heights, New York.
Oshani Seneviratne is the director of health data research
at the Institute for Data Exploration and Applications at the
Rensselaer Polytechnic Institute.