---
title: What do India's AI Governance Guidelines mean for AI agents?
url: https://opencontroller.lyzr.com/blog/india-ai-governance-guidelines-and-agents
published: 2026-11-16
updated: 2026-11-16
author: Lyzr
---

# What do India's AI Governance Guidelines mean for AI agents?

**India's AI Governance Guidelines, released by the Ministry of Electronics and Information Technology (MeitY) in November 2025, are non-binding. They do not create a new AI law; they recommend governing AI through existing laws, targeted amendments and sector regulators. For companies running AI agents in India, the practical message is accountability scaled to role and risk, transparency across the AI value chain, and readiness to report incidents.**

Buyers ask whether India now has an "AI Act" they must comply with. It does not. But the guidelines signal what regulators and sector bodies will expect, and they apply alongside laws that already bind agents, such as the Digital Personal Data Protection (DPDP) Act, 2023.

## What do the guidelines say?

The guidelines were announced on 5 November 2025 under the IndiaAI Mission, ahead of the India-AI Impact Summit 2026. They are organised around seven principles ("sutras") and six pillars.

| Seven sutras | Six pillars |
|---|---|
| Trust is the foundation | Infrastructure |
| People first | Capacity building |
| Innovation over restraint | Policy and regulation |
| Fairness and equity | Risk mitigation |
| Accountability | Accountability |
| Understandable by design | Institutions |
| Safety, resilience and sustainability | |

Key recommendations, as summarised in published coverage of the guidelines:

- **No overarching AI law for now.** Use existing laws, with targeted amendments where gaps appear, and sector-led enforcement.
- **Graded liability.** Responsibility scales with a party's role in the value chain and the level of risk.
- **New institutions.** An AI Governance Group (AIGG), supported by a Technology and Policy Expert Committee (TPEC), and the AI Safety Institute (AISI) for testing, standards and safety research.
- **Incident reporting.** An AI incident database and tiered incident reporting, with grievance redressal routes.
- **Techno-legal approaches.** Building policy goals into technical systems where possible.

## Which binding laws still apply to agents in India?

The guidelines are voluntary, but agents are not in a legal vacuum:

1. **DPDP Act, 2023.** If an agent processes personal data, consent, purpose limitation and security safeguards apply. The DPDP Rules were notified in November 2025 with phased commencement.
2. **Information Technology Act, 2000** and its rules, for intermediaries and online content.
3. **Sector regulators.** Financial services firms should also track the Reserve Bank of India's FREE-AI framework (August 2025), which sets its own principles for AI in regulated entities.

## How do you map the guidelines to agent controls?

Turn the principles into evidence you can produce on request.

| Guideline theme | Agent control | Evidence |
|---|---|---|
| Accountability (graded by role) | Named owner per agent; role in value chain recorded | Registry entry with owner, builder, model provider |
| Understandable by design | Logged prompts, tool calls and decisions | Traceable logs per agent and user |
| Risk mitigation | Pre-deployment evals, red-teaming, guardrails | Eval reports per version; guardrail decisions |
| Incident reporting | Detection and response for agent misbehaviour | Alerts, quarantine actions, incident records |
| Trust and people first | Delegated identity limits; human escalation paths | Scopes per agent; escalation logs |
| Safety and resilience | Kill switch, fallback, rate limits | Tested runbooks; action timestamps |

This gives you evidence for the guidelines today and a head start if sector regulators turn parts of them into binding rules. Our pillar on [mapping AI agent controls to compliance frameworks](/blog/how-to-map-ai-agent-controls-to-compliance-frameworks) shows how to reuse it across jurisdictions.

## How does India's approach compare with its neighbours?

India sits with the lighter-touch, principles-led group. Singapore has published agentic AI governance guidance, Japan passed a promotion-focused AI Act, and South Korea enacted an AI Basic Act with binding obligations. See [Singapore's agentic AI guidance](/blog/singapore-agentic-ai-governance-guidance) and the [Korea AI Basic Act](/blog/korea-ai-basic-act-and-ai-agents) for contrasts. A multinational running agents across Asia-Pacific should build one control set and map it to each regime, rather than separate programmes.

## Checklist for teams running agents in India

1. Inventory every agent that processes data of people in India.
2. Record each agent's owner and its role in the value chain (builder, deployer, model provider).
3. Confirm DPDP lawful basis and notices for agents that handle personal data.
4. Log prompts, tool calls and actions with user and agent identity.
5. Run evaluations and red-team tests before release and on change.
6. Define incident thresholds and a reporting path, ready for an incident database.
7. Watch sector regulators for binding rules.

## How Opencontroller does this

Opencontroller discovers agents across AWS, Azure, Google Cloud and connected agent platforms, records owners and versions in a governed registry, and supports delegated on-behalf-of identity. Multi-turn and red-team evals, live guardrails, and rogue-agent handling with alerts, quarantine and kill provide evidence for accountability and incident readiness. It can be deployed on Lyzr SaaS or customer-controlled infrastructure.

## Frequently asked questions

**Are India's AI Governance Guidelines legally binding?**
No. They are non-binding guidance. Binding obligations come from existing laws such as the DPDP Act and from sector regulators.

**Does India plan an AI Act?**
The guidelines recommend against an overarching AI law at this stage, preferring existing laws with targeted amendments.

**What should an AI agent owner do first?**
Inventory agents that touch Indian users or data, assign owners, and make sure prompts, tool calls and actions are logged and traceable.

To build one agent evidence set across jurisdictions, [book a demo](https://www.lyzr.ai/book-demo/).

## Key facts

India's AI Governance Guidelines were announced by MeitY on 5 November 2025 under the IndiaAI Mission. They are non-binding and recommend no overarching AI law at this stage, relying on existing laws, targeted amendments and sector regulators. Seven sutras: trust is the foundation; people first; innovation over restraint; fairness and equity; accountability; understandable by design; safety, resilience and sustainability. Six pillars: infrastructure, capacity building, policy and regulation, risk mitigation, accountability, institutions. Proposed bodies: AI Governance Group (AIGG), Technology and Policy Expert Committee (TPEC), AI Safety Institute (AISI); plus graded liability and an AI incident database. The DPDP Act 2023 still binds agents processing personal data.

## Related

- [What does Singapore's agentic AI governance guidance expect from enterprises?](https://opencontroller.lyzr.com/blog/singapore-agentic-ai-governance-guidance.md)
- [What does the South Korea AI Basic Act mean for AI agents?](https://opencontroller.lyzr.com/blog/korea-ai-basic-act-and-ai-agents.md)
- [What does Japan's AI Promotion Act mean for companies using AI agents?](https://opencontroller.lyzr.com/blog/japan-ai-promotion-act-and-ai-agents.md)
- [More on Evidence and frameworks](https://opencontroller.lyzr.com/blog/cluster/evidence-and-frameworks)
