"India has no AI law, so we can do what we like" is one of the more expensive misunderstandings in Indian business today. It is true that India has no single, dedicated Artificial Intelligence Act. It is not true that AI is unregulated. AI systems process personal data, publish content, make decisions about customers and run on IT infrastructure, and each of those activities is already covered by Indian law and regulator guidance.
This article maps the AI laws in India that apply today, with official or reputable sources, and ends with a compliance starter list. It is a practical overview, not legal advice.
Is there a dedicated AI law in India?
Not at the time of writing. The government's approach has been to govern AI through existing laws plus targeted rules and guidance. In November 2025, MeitY unveiled the India AI Governance Guidelines under the IndiaAI Mission. As reported by News On AIR, Principal Scientific Adviser Prof. Ajay Kumar Sood described "Do No Harm" as the framework's guiding principle, with sandboxes for innovation and risk mitigation in a flexible, adaptive system. The guidelines are a governance framework, not a statute, but they signal how regulators will think about AI risk.
The laws that already govern AI use in India
1. Digital Personal Data Protection Act, 2023 and DPDP Rules, 2025
If your AI system touches personal data, whether training data, customer chats, resumes, call recordings or CCTV, the DPDP Act applies. The text of the Act requires a lawful basis such as consent for a specified purpose, reasonable security safeguards, and a way to honour the rights of data principals. As Bar & Bench explains, failure to take reasonable security safeguards under Section 8(5) can attract a penalty of up to ₹250 crore, Rule 6 of the DPDP Rules 2025 lists minimum security measures, and substantive compliance with Section 8(5) begins in May 2027. For AI, this means asking: what personal data goes into the model or prompt, where it is processed, who can see it, and how long it is kept.
2. Information Technology Act, 2000 and the IT Rules, 2021
The IT Act and the Intermediary Guidelines and Digital Media Ethics Code Rules govern platforms that host or transmit user content, including AI-generated content.
The March 2024 MeitY advisory. MeitY advised intermediaries and platforms that their AI models and tools must not allow users to create or share unlawful content, must not permit bias or threaten electoral integrity, and that under-tested or unreliable AI models should be offered in India only with clear labelling of their possible fallibility, according to News On AIR.
The 2026 amendment on synthetic content. MeitY notified amendments to the IT Rules on 10 February 2026, effective 20 February 2026. As summarised by Freshfields, they define "synthetically generated information" as audio, visual or audio-visual content that appears real, require such content to be clearly and prominently labelled, stop labels and metadata from being removed, and shorten takedown windows: three hours for court or government orders and two hours for certain high-risk content such as impersonation and non-consensual intimate imagery. Significant social media intermediaries must ask users to declare whether uploads are synthetic and use technical measures to verify those declarations. Intermediaries that fail these due diligence duties risk losing safe-harbour protection.
If your product lets users generate or share images, audio or video, these rules are directly relevant. If you use AI to produce marketing videos or voice-overs, expect platforms to require labels.
3. CERT-In directions on cyber incidents
CERT-In's directions of 28 April 2022, issued under Section 70B(6) of the IT Act, cover information security practices and the reporting of cyber incidents. An AI system is part of your IT estate: a data leak through a chatbot, a compromised AI API key or an agent misused to access systems can be a reportable incident. Include AI services in your incident response plan and logging.
4. Sector regulators
Regulated sectors carry additional expectations. The clearest example is the Reserve Bank of India's FREE-AI Committee report of 13 August 2025, the Framework for Responsible and Ethical Enablement of Artificial Intelligence. It sets out seven "sutras", including Trust is the Foundation, People First, Fairness and Equity, Accountability and Understandable by Design, with 26 recommendations. Among its themes: board-approved AI policies for regulated entities, stronger incident reporting, extending product approval, consumer protection and audit processes to AI, and telling consumers when they are interacting with AI. Banks, NBFCs and fintechs should treat it as the direction of travel even where specific rules are still being written. In our view, companies regulated in securities, insurance or healthcare should assume their existing duties on fairness, record-keeping and outsourcing extend to AI systems too.
5. General laws that apply to AI outputs
- Consumer protection: misleading claims made by an AI chatbot or in AI-generated ads are still misleading claims by your business.
- Copyright: AI-generated content that reproduces protected work can infringe, and ownership of AI-generated work is still developing.
- Contract and negligence: if your AI gives wrong advice that a customer relies on, the usual liability principles apply.
- Employment and anti-discrimination principles: AI used in hiring or performance decisions must not produce unfair outcomes you could not defend if a person had made them.
How the pieces fit together
| Your AI activity | Main Indian rules to check |
|---|---|
| Chatbot handling customer queries | DPDP Act (personal data in chats), consumer protection, sector guidance if regulated |
| Platform where users generate images, audio or video | IT Rules 2021 as amended in 2026 (labelling, takedowns), March 2024 advisory |
| AI screening resumes or scoring customers | DPDP Act, fairness expectations, sector rules such as RBI guidance for credit |
| AI agents with access to internal systems | CERT-In directions (incidents), DPDP security safeguards |
| Publishing AI-generated marketing content | Copyright, consumer protection, platform labelling rules |
| Selling AI products to EU clients | All of the above plus the EU AI Act |
AI compliance starter list for Indian companies
- Make an AI inventory. Every AI tool, model, API and feature in use, who owns it and what data it touches.
- Map personal data flows for each AI use, including prompts, logs and vendor processing locations.
- Write an AI-use policy covering approved tools, data classes that may and may not be used, and human review requirements.
- Label AI interactions and synthetic media where users could be misled.
- Add AI to security and incident response: access controls, logging, key management and CERT-In reporting paths.
- Review vendor contracts for data use, retention, security, location and liability.
- Keep humans accountable for decisions about people: credit, hiring, claims, discipline.
- Track sector guidance, especially if you are regulated by RBI, SEBI or IRDAI.
- Review every six months. This area is changing quickly; the 2026 amendments arrived with very short notice.
How Cehpoint can help
Our AI readiness assessment produces an AI roadmap for your business; the inventory and data mapping in the starter list above are natural inputs to it. Where personal data is central, our DPDP compliance service helps you put the Act's duties into practice. Companies selling AI into Europe can add our EU AI Act compliance work.