National Strategy for AI, 2018 (NITI Aayog)

Short answer

NITI Aayog wrote this paper after the 2018–19 Budget speech asked it to set up a national AI programme (Introduction, PDF page 5). The paper calls itself a discussion paper (PDF page 5). It names the brand "#AIforAll". It says AI should help growth reach everyone.

Law: none. Nothing in this paper binds a company or a person.

What it means for you

You do not need to follow this paper. Read it to see how India first framed AI. Later papers, such as Responsible AI for All and the AI Governance Guidelines, build on its ideas. If you sell AI to government, its focus sectors may tell you where public money and projects go.

How it works

Five focus sectors (Executive Summary, PDF page 7):

• Healthcare — better access to quality care at lower cost.
• Agriculture — higher farm income and productivity, and less waste.
• Education — better access and quality.
• Smart cities and infrastructure — serving a growing urban population.
• Smart mobility and transportation — smarter and safer travel.

Five barriers the paper wants fixed (PDF page 7):

• Too few people skilled in AI research and use.
• No good systems for sharing useful data.
• High cost and low awareness of AI.
• Privacy and security, including no formal rules on anonymising data.
• Too little teamwork on adopting AI.

Two kinds of research centre (PDF page 7). Centres of Research Excellence (COREs) would do core research. International Centres of Transformational AI (ICTAIs) would build and deploy applied AI, with private-sector partners. The paper explains COREs further on PDF page 54.

A shared AI cloud. The paper proposes AIRAWAT, a cloud platform for big data and AI computing (PDF page 59). This was a proposal in the paper.

Privacy. The paper notes Europe's GDPR as an example of rules that push for less privacy-invasive design (PDF page 87). GDPR is EU law. It is not Indian law. India's own data law today is the DPDP Act, 2023.

Where it fits

India's AI papers come in layers. This strategy came first and set the goals. Responsible AI for All then set principles for how AI should behave. The IndiaAI Mission put money behind computing, data and skills. The AI Governance Guidelines, 2025 now give the main advice for companies. None of these is a law. Your legal duties come from laws such as the DPDP Act, 2023, the IT Act, 2000 and the IT Rules, 2021.

Good practice: if you pitch an AI project to a public body, link it to one of the five sectors above. Explain how it deals with the barriers the paper names, such as data access and privacy.

What's still open

• The paper does not say which of its ideas the government later funded. We do not say either.
• Good practice: treat it as background reading, not a compliance guide.

Next: Responsible AI for All (NITI Aayog, 2021) and the AI Governance Guidelines, 2025.