Dr.Rupali Singh, Director- LRC, Atmiya University, Rajkot, Gujarat (India)
Saanvi Kumar, BA.LLB (Hons.), Symbiosis Law School, Pune, Maharashtra (India)
Think of a farmer in rural Bihar using an AI app that can predict crop yields in Hindi, or a student in Kerala getting personalized tutoring that spots biases before they can harm. Now, imagine these tools not just innovative but also safe—free from deepfakes, discrimination, or unchecked surveillance. This is not science fiction; it is the promise of India’s newly released AI Governance Guidelines, published by MeitY on November 5, 2025. As the world grapples with AI’s double-edged sword—driving $15 trillion in global value by 2030 while risking misinformation and inequality—India is stepping up with a practical playbook. Anchored in Prime Minister Narendra Modi’s “AI for All” vision, these guidelines aim at powering inclusive growth toward Viksit Bharat by 2047 without strangling startups or overburdening developers.
Drafted by a high-powered committee chaired by IIT Madras’ Prof. Balaraman Ravindran, with inputs from over 2,500 public submissions, the 68-page document revises an earlier draft to emphasize “techno-legal” solutions—blending code, policy, and accountability. Backed by the ₹10,000 crore IndiaAI Mission, it invests in GPUs, data platforms, and skilling to democratize AI. But what sets it apart? A “light-touch” approach: voluntary measures first, heavy regulation only where risks demand it. As Principal Scientific Adviser Prof. Ajay Sood notes in the preface, “AI must serve the common good, ensuring benefits reach the last citizen.”
From UPI Success to AI Safeguards
AI is not new-it has been in development since the 1950s-but the latest generative models, such as ChatGPT, turbocharged it, allowing systems to reason, create, and act virtually autonomously. In fact, for a country like India with 1.4 billion people and a thriving digital economy, AI potentially supercharges healthcare diagnostics, multilingual education on Bhashini, or climate-resilient farming. Yet, it brings equally large risks: algorithms amplify caste divisions, deepfakes undermine trust in elections, or unchecked surveillance via DPDPA.
The AI Act bans high-risk uses strictly at the global scale, while Washington banks on voluntary codes. India navigates a middle path: scalable and inclusive AI riding on Digital Public Infrastructure such as Aadhaar and UPI. The guidelines basically build on the work done by RBI’s FREE-AI report and adopt its principles for all sectors. As MeitY Secretary S. Krishnan writes in the foreword, “Without robust trust and safety, our infrastructure efforts risk stalling adoption.” The release is quite timely amidst the buzz in the India AI Summit-aligned events, NISAR satellite launches, and IAF’s AI integrations.
The Seven ‘Sutras’: Guiding Principles for Ethical AI
At the heart are seven “sutras” (guiding principles), technology-agnostic beacons for developers, regulators, and users. Adapted from RBI’s framework and aligned with national priorities, they prioritize humans over hype, Trust is the Foundation: Without it, innovation stagnates. Mandate transparency in AI decisions to build public confidence—think labelling deepfakes with 10% content tagging, People First: Human-centric design with oversight and empowerment. AI should augment, not replace—e.g., doctors retaining final say in rural diagnostics, Innovation over Restraint: Favor responsible progress over caution. Sandboxes for testing let startups like Sarvam AI iterate freely , Fairness & Equity: Promote inclusion, avoid discrimination. Address biases in datasets to ensure AI serves marginalized groups, from SC/ST communities to women in fintech , Accountability: Clear responsibility chains. A graded liability system holds deployers accountable based on risk and due diligence , Understandable by Design: Disclosures in plain language—for users and regulators. Explain AI outputs like “This recommendation is based on X data, with Y confidence.” , Safety, Resilience & Sustainability: Robust systems against shocks, eco-friendly (e.g., energy-efficient models). Protect kids from addictive AI via engagement limits.
These aren’t abstract; they’re actionable, echoing the Future of Privacy Forum’s call for DPDPA-AI synergy.
Roadmap: Short, Medium, and Long-Term Actions
No vision without execution. Part 3’s action plan timelines progress:
- Short-term (0-1 year): Launch AISI, voluntary codes, awareness drives; expand GPU access.
- Medium-term (1-3 years): Publish standards, regulatory sandboxes; pilot DPI-AI integrations like UPI for AI payments.
- Long-term (3+ years): Draft new laws on emerging risks (e.g., agentic AI); review ecosystems for sustainability.
Institutions map roles: MeitY leads, NITI Aayog advises, BIS sets standards. As iSPIRT’s Sharad Sharma (committee member) tweeted, it’s “law, code, and accountability converging.”
Early Reactions: Optimism with Cautious Notes
The launch has created a buzz. On X, ProductNation praised “techno-legal as foundational,” linking to DEPA for ethical data. Dhiraj Goswami termed it “a game-changer for global innovation, eyeing $1T GDP boost.” An editorial in Hindustan Times praised the “light-touch” balance.
Yet concerns remain. Anvar Sadath from Kerala warned that “surveillance risks” were imminent with voluntary rules and Centre-led oversight, mirroring the debates over IT Rules. The Hindu noted a “hands-off” approach, different from earlier drafts, cautioning against falling foul of privacy. All in all-true to the cause: No new laws, only nimble evolution.
Part 4 offers practical tips: Industry—comply with laws, publish reports, use grievance mechanisms. Regulators—avoid compliance traps, embrace techno-legal innovation.
Paving the Path: India’s AI Governance in Action
These are not red tape but rocket fuel for ethical AI. By putting “people first” and innovation, India is set to be an exporter of trusted tech globally, from multilingual models to bias-free tools. For startups, green lights come via sandboxes; for citizens, safeguards against harm.
As we hit the India AI Impact Summit 2026, the real test indeed begins: Will AISI actually deliver robust testing? Can DPI scale AI equitably? Engage now: download guidelines, participate in consultations, or build that next app. Which sutra resonates most with you? Share in the comments. India’s AI story is only just beginning, and it is one that the world will be watching.
