NITI Aayog Discussion Paper: An Aspirational Step towards India's AI Policy

NITI Aayog Discussion Paper: An Aspirational Step towards India’s AI Policy is a policy commentary authored by Sunil Abraham, Elonnai Hickok, Amber Sinha, Swaraj Barooah, Shweta Mohandas, Pranav M. Bidare, Swagam Dasgupta, Vishnu Ramachandran, and Senthil Kumar at the Centre for Internet and Society (CIS).

The brief analyses NITI Aayog’s 2018 discussion paper on India’s National Strategy for Artificial Intelligence, assessing its ambitions, policy directions, and critical omissions. It commends the initiative as a significant step toward a national AI strategy while identifying areas where the paper falls short — particularly on issues of privacy, ethics, fairness, accessibility, and accountability.

Publication Details
Abstract
Context and Background
Key Themes or Findings
Full Text
Citation

Publication Details

👤 Authors:: Sunil Abraham, Elonnai Hickok, Amber Sinha, Swaraj Barooah, Shweta Mohandas, Pranav M. Bidare, Swagam Dasgupta, Vishnu Ramachandran, and Senthil Kumar
🏛️ Published by:: The Centre for Internet and Society (CIS)
📅 Date:: June 2018
📘 Type:: Policy Analysis / Commentary on National AI Strategy
📄 Access:: Download PDF

Abstract

This commentary critically examines NITI Aayog’s “National Strategy for Artificial Intelligence” (2018), the first formal step toward a national AI policy in India. The report identifies the discussion paper’s positive focus on AI’s potential in healthcare, agriculture, education, smart cities, and mobility, while pointing out several gaps in inclusion, accessibility, and governance.

The authors argue that the policy, while aspirational, overlooks key sectors like manufacturing, services, and defence, and fails to recognise the transformative potential of AI for persons with disabilities. They also caution that AI-enabled surveillance in smart cities risks violating constitutional rights and privacy norms.

Overall, the analysis calls for an AI strategy grounded in transparency, fairness, accountability, and human rights, guided by open standards, free and open source software (FOSS), and strong ethical oversight.

Context and Background

The NITI Aayog discussion paper is India’s first major government document outlining an AI roadmap. It recognises AI’s transformative potential across public and private sectors but adopts a largely techno-optimistic and industry-driven lens.

While it highlights five social-impact areas—healthcare, agriculture, education, smart cities, and mobility—it omits vital economic sectors such as manufacturing and services, which are central to employment and safety. Equally absent are discussions on defence, despite India’s international leadership on lethal autonomous weapons (LAWS), and on AI for accessibility, despite prior recognition in the Task Force Report on Artificial Intelligence (2018).

The paper’s endorsement of AI-driven surveillance for “public safety” in smart cities raises legal and constitutional red flags. It conflicts with due process principles and threatens fundamental rights like privacy, freedom of expression, and freedom of assembly.

The commentary therefore calls for a rights-based AI policy that aligns India’s strategy with international human rights standards and emphasises legality, legitimacy, necessity, and proportionality in all AI use cases.

Key Themes or Findings

Focus Areas and Omissions:
The five identified areas are important, but excluding manufacturing, services, defence, and accessibility limits the policy’s inclusivity.
Research and Intellectual Property:
The paper proposes Centres of Research Excellence (COREs) and International Centres for Transformational AI (ICTAIs) but ignores open research principles such as FOSS, open data, and open standards.
It inaccurately portrays India’s patent regime as a barrier, failing to appreciate Section 3(k) of the Patents Act, which prohibits software patents—a safeguard for innovation and access.
Ethics, Fairness, and Transparency:
Although the discussion paper raises ethical concerns, it lacks concrete solutions. CIS recommends:
- Mandatory auditability of government AI systems;
- Recognition of a right to reverse engineer under the Copyright Act;
- Public availability of source code for rights-impacting AI;
- Regulatory tools like bias audits, discrimination impact assessments, and certification mechanisms to ensure fairness.
Privacy and Data Protection:
The paper acknowledges privacy but omits key data rights such as right to explanation, opt-out of automated processing, and data minimisation. It also fails to integrate international frameworks like the IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems.
AI Regulation and Sectoral Standards:
Over-reliance on industry self-regulation weakens accountability. CIS recommends a co-regulatory framework, combining industry codes with government oversight, aligned with the UN Guiding Principles on Business and Human Rights.
Security, Safety, and Accountability:
The policy should mandate:
- Kill switches for all kinetic AI systems;
- Human-in-the-loop for safety-critical AI applications;
- Human rights impact assessments before deployment;
- Legal protection for whistleblowers and security researchers.
Education and Research:
Ethical and human rights considerations must be integrated into AI curricula across educational institutions. CIS recommends a national AI ethics curriculum and corporate-funded AI literacy programmes under government supervision.
Data Standards and Sharing:
The policy’s proposal for corporate and government data sharing must include privacy-enhancing technologies like anonymisation, pseudonymisation, and differential privacy, along with anti-monopoly provisions to prevent data hoarding.
India as an AI Test-Bed:
The authors caution against treating India as an “AI garage” or unregulated experimental ground for emerging economies, stressing the need for ethical governance and informed consent before testing high-risk technologies.

In summary, the NITI Aayog paper is an ambitious and valuable milestone — but to build an inclusive, ethical, and rights-respecting AI ecosystem, India’s AI strategy must evolve beyond aspiration to accountability.

Full Text

📄 Download Full Report (PDF)

Citation

If you wish to reference or cite this publication, you may use one of the following formats.

APA

Abraham, S., Hickok, E., Sinha, A., Barooah, S., Mohandas, S., Bidare, P. M., Dasgupta, S., Ramachandran, V., & Kumar, S. (2018).
NITI Aayog Discussion Paper: An Aspirational Step towards India's AI Policy.
The Centre for Internet and Society (CIS).
https://sunilabraham.in/publications/niti-aayog-discussion-paper-ai-policy/

BibTeX

@misc{abraham2018niti,
  author = {Abraham, Sunil and Hickok, Elonnai and Sinha, Amber and Barooah, Swaraj and Mohandas, Shweta and Bidare, Pranav M. and Dasgupta, Swagam and Ramachandran, Vishnu and Kumar, Senthil},
  title = {NITI Aayog Discussion Paper: An Aspirational Step towards India's AI Policy},
  year = {2018},
  howpublished = {The Centre for Internet and Society (CIS)},
  url = {https://sunilabraham.in/publications/niti-aayog-discussion-paper-ai-policy/}
}

MLA

Abraham, Sunil, Elonnai Hickok, Amber Sinha, Swaraj Barooah, Shweta Mohandas, Pranav M. Bidare, Swagam Dasgupta, Vishnu Ramachandran, and Senthil Kumar.
"NITI Aayog Discussion Paper: An Aspirational Step towards India's AI Policy."
The Centre for Internet and Society (CIS), June 2018.
https://sunilabraham.in/publications/niti-aayog-discussion-paper-ai-policy/