Government AI Policy in India: The Complete 2026 Guide for the AI Industry

When it comes to building, investing in, and implementing AI in India, there has been quite a bit of change in the way things work for you. The country has gone from a strategic document in 2018 to a national mission with funding and a governance framework published in 2026.

How India’s AI Policy Got Here

India’s strategy was not an overnight realization. It developed over eight years through three separate efforts.

First, it began in 2018 when the NITI Aayog issued the National Strategy for Artificial Intelligence. This established the initial vision of the country through the #AIforAll initiative.

Six years later, in March 2024, the IndiaAI Mission with an allocation of ₹10,372 crores received approval from the Cabinet. This funding initiative enabled the construction of the compute, data, and skills infrastructure.

Finally, in November 2025, the India AI Governance Guidelines were released followed by a full-scale launch at the AI Impact Summit in February 2026. This completed the third part of the puzzle – guidelines for developing and governing the use of AI.

The process was intentional. Fund the infrastructure first. Govern the developments next.

The Three Layers of India’s AI Policy

The government’s policy on AI for India consists of three interrelated tiers, with each having its own owner.

1. The first tier is finance and infrastructure, which is provided through the IndiaAI Mission and owned by MeitY. 
2. The second is the governance and risk management, which is provided through the 2026 Guidelines. 
3. The third is the original sectoral strategy that was created in 2018 by NITI Aayog and continues to define the sectors of public investment into AI.

These tiers are not separate from each other but should work together as a single framework.

The India AI Mission: What It Actually Funds

IndiaAI Mission is the financial muscle powering India’s ambitions in artificial intelligence. It consists of seven financial pillars:

• Compute Infrastructure – AI Infrastructure shared via public-private partnerships. India’s national compute infrastructure passed the 34,000 GPUs mark by mid-2025 and has increased since.
• Innovation Centre – indigenously created large multimodal and domain-specific foundational models.
• Dataset Platform – simplified access to quality non-personal datasets.
• AI Solutions Scaling Initiative – scaling AI solutions to achieve socio-economic impact.
• FutureSkills – AI Education initiatives and creation of Data and AI labs in tier 2 and tier 3 cities.
• Startup Financing – funding for deep tech AI startups.
• Safe & Trusted AI – indigenous technology and governance structure for safe AI deployment.

As a founder or researcher who needs grants, compute resources, or datasets, this is the layer you care about the most.

The 2026 Governance Guidelines: Seven Guiding Principles

Whereas The Mission is where the funding of AI is done, The Guidelines are there to oversee it. This framework is based on the seven “sutras” which the drafting committee says should guide it.

1. Trust is the Foundation. Trusting the AI is about incorporating trust within the AI technology, the companies developing it, and those regulating it.
2. People First is about ensuring that humans remain in control in all aspects and AI technology only augments their decisions but does not replace them.
3. Innovation over restraint is about prioritizing responsible innovation over restraint of blocking whenever risks are being managed.
4. Fairness and equity involve designing and testing systems such that biases do not arise, especially those that target marginalized individuals.
5. Accountability involves remaining visible and taking responsibility for what the systems are doing.
6. Design for Explainability is about incorporating explainability into the system right from the onset.
7. Safety, resilience, and sustainability involves making sure the systems have high resilience, can detect any anomalies, and are sustainable and resource efficient.

None of these principles are legal provisions per se. They just guide the application of law on AI.

Six Policy Pillars Behind the Guidelines

Below the umbrella of the above seven principles, the Guidelines structure their recommendations into six pillars of action.

1. Infrastructure refers to increasing compute capacity and data access, along with integration of AI with Digital Public Infrastructure such as Aadhaar, UPI, and DigiLocker.
2. Capacity Building relates to capacity development within the government, law enforcement, and citizens, and the development of AI skills in smaller cities.
3. Policy & Regulation is concerned with the evaluation of existing laws in relation to AI and sandboxing of new regulation.
4. Risk Mitigation refers to the creation of AI risk classification specifically relevant to India, and the national incident reporting framework.
5. Accountability includes clarification of the applicability of current laws on AI and mandatory complaint filing options.
6. Institutions relate to the establishment of the new institutional bodies mentioned in the next section.

Laws That Already Apply to AI in India

There is an existing misconception about AI regulation in India being in a legal vacuum, until new laws emerge. It is not entirely accurate. There are existing laws that apply.

• IT Act, 2000 acts as a basic regulatory law covering digital intermediaries, cybercrime, and platform liability, including AI enabled platforms.
• IT Rules, 2021, in addition to 2026 amendments, act as the base for intermediary liability rules, while 2026 amendments provide special provisions on AI and deepfakes.
• Digital Personal Data Protection Act, 2023 regulates the issues related to consent, data processing, and fiduciary obligations of any AI interacting with personal data.

All of the above is supplemented by sector specific regulations imposed by such regulatory bodies as RBI, SEBI, and IRDAI in their respective sectors.

New Institutions Being Built

For this, the Guidelines recommend that there be set up three new or re-constituted organizations.

• The AI Governance Group (AIGG) is to coordinate the development of all policies and ensure that AI governance remains in line with national objectives.
• The Technology and Policy Expert Committee (TPEC) is to provide expert technical advice to the AI Governance Group on significant national/international issues of AI.
• AI Safety Institute (AISI) will conduct safety-related research, develop evaluation standards and benchmarks and advise both regulatory agencies and industry.

These organizations complement existing ones such as MeitY, CERT-In, and the Data Protection Board of India. They do not replace them.

A Practical Compliance Checklist

If you are developing or operating AI systems in India right now, then these are the Guidelines’ actual expectations of you.

1. Comply with the current sectoral legislation. Data privacy, copyright, consumer protection laws, and those protecting women and children already apply to your AI.
2. Be ready to provide evidence for that compliance. Regulators may request it rather than your assurances.
3. Implement voluntarily the currently available safeguards. Privacy measures, fairness assessments, and bias analyses are required even if there are no mandatory requirements yet.
4. Establish a complaint mechanism. There should be a way for people to submit their grievances about harms related to AI, with a reasonable time limit for resolution.
5. Submit transparency reports. Assess and disclose the risks of harm associated with your AI in the context of India. Confidential reports can be submitted to regulators rather than public ones.
6. Investigate techno-legal solutions. Privacy preserving techniques, machine unlearning, algorithm auditing, and automated bias assessment are among the required actions.

No “AI license” needs to be obtained. The required action is diligence demonstrated constantly.

The Action Plan, in Three Phases

Rollout in India is staged and not immediate. Over the short term, efforts will center around establishment of the AIGG and TPEC, development of risk classification frameworks for India, conducting a regulatory gap analysis, and releasing a master circular detailing applicable regulations.

Over the medium term, focus will move towards developing common standards related to content verification and cybersecurity, creation of an AI incident database in the country, amending applicable laws in case of any confirmed gaps, and piloting of regulatory sandboxes in high risk sectors.

Over the long term, efforts will continue to review the entire regulatory framework, develop new laws for emerging risks, increased international collaboration regarding AI standards, and horizon scanning for future considerations.

Where This Started: The 2018 #AIforAll Strategy

In the absence of all that, NITI Aayog’s strategy of 2018 set the initial path for India. It recognized five sectors which had the maximum social value of AI applications.

• Healthcare – Early diagnosis, personalized care and enhanced access in underprivileged areas.
• Agriculture – Prediction of crop yields, pest detection and advisories for farmers in real time.
• Education – Learning technologies tailored to individual needs and identification of students at risk.
• Smart Cities and Smart Infrastructure – Traffic Management, Utility monitoring and public safety.
• Smart Mobility and Transportation – Congestion prevention, railways safety and autonomous driving.

A lot of the current priorities of the IndiaAI Mission can be traced back to the 2018 list, although the funding mechanism and governance associated with it is all new now.

How Government Itself Uses AI

Different from the ways in which India regulates artificial intelligence companies, another issue lies in how the Indian government itself uses AI technology. This is what the OECD’s 2025 report called Governing with Artificial Intelligence addresses.

This report describes about 200 cases of AI application in eleven areas related to government functions – public services, justice, anticorruption, public finances, and others. The main idea of this report is that there are certain measures that must be taken into consideration with regard to the application of AI technologies in the field of governance, and these are different from how it regulates the industry.

The introduction of artificial intelligence into Digital Public Infrastructure, which is included in the Guidelines’ infrastructure pillar, is one of those trends.

How India Compares to Other Countries

The Indian approach is based on the principles of voluntariness and “innovation over restraint,” making it more similar to the American one that uses light-handed enforcement through executive orders in a sector-specific manner.

In turn, the EU AI Act represents a different strategy as it entails mandatory compliance with obligations depending on the risk category into which an AI application falls.

The Indian Guidelines are more evidence-based than the other two and opt for regulation rather than a license at the outset; however, they are more centralized than the other two strategies due to the newly established AIGG and TPEC framework.

Frequently Asked Questions

Key Takeaways

The policy for AI in India comprises the IndiaAI mission, the 2026 guidelines for AI, and the 2018 NITI Aayog strategy for AI. There are currently laws that cover AI, while there are newly created bodies to deal with new challenges. The thrust is on innovating with responsible governance.