New Delhi, Wednesday, November 5, 2025 – The Ministry of Electronics and Information Technology (MeitY) unveiled the India AI governance guidelines today. Officials released the document during an event chaired by Principal Scientific Adviser Ajay Kumar Sood. The guidelines aim to foster responsible AI adoption across sectors while mitigating risks like bias and misinformation.
These India AI governance guidelines matter deeply for South Asia. India leads the region’s digital economy, with AI powering cross-border services in finance, health, and agriculture. Neighbours like Pakistan, Bangladesh, and Sri Lanka rely on Indian cloud infrastructure for data processing. The framework promotes standards that could harmonise regulations, easing data flows and reducing risks from unregulated AI imports. It positions India as a model for Global South nations balancing growth with equity.
The guidelines stem from the IndiaAI Mission, approved in 2024 with INR 10,372 crore funding. A subcommittee under Prof. B. Ravindran of IIT Madras drafted the initial report in January 2025. Public consultations drew over 650 inputs, leading to refinements after 20 meetings. The final version emphasises self-regulation over heavy-handed laws.
MeitY AI Governance Framework India: Core Structure
The MeitY AI governance framework India adopts a techno-legal approach. It embeds compliance into AI systems from design to deployment. Key elements include a risk-based classification for AI tools. Low-risk applications face minimal oversight. High-risk ones in critical sectors like healthcare and justice require audits and transparency reports.
The framework proposes three institutions for oversight. The AI Governance Group (AIGG) coordinates policy. The Technology and Policy Expert Committee (TPEC) handles assessments. The AI Safety Institute (AISI) sets technical standards. Developers must disclose model limitations and conduct impact assessments.
MeitY Secretary S. Krishnan stated: “Our focus remains on using existing legislation wherever possible. At the heart of it all is human centricity, ensuring AI serves humanity and benefits people’s lives while addressing potential harms.” This aligns with the Information Technology Act and Digital Personal Data Protection Act.
An action plan outlines timelines. Short-term steps cover awareness campaigns and risk frameworks by mid-2026. Medium-term goals include incident reporting systems and sandboxes by 2027. Long-term measures build liability regimes and international alignments.
India Safe Trusted AI Guidelines: Risk and Accountability Focus
The India safe trusted AI guidelines prioritise harm prevention. AI systems must undergo bias testing before launch. Platforms face graded liability: developers for design flaws, deployers for misuse. Intermediaries report incidents within 24 hours.
The guidelines ban high-risk uses like social scoring without oversight. They mandate watermarking for synthetic content, extending the 2024 deepfake advisory. Consent remains central for data training, with opt-out rights for users.
Prof. Ajay Kumar Sood, Principal Scientific Adviser, remarked: “The guiding principle that defines the spirit of the framework is simple, ‘Do No Harm’. We focus on creating sandboxes for innovation and on ensuring risk mitigation within a flexible, adaptive system.” This ethos supports the mission’s safe AI pillar, allocated INR 600 crore.
Industry must adopt voluntary codes for transparency. Regulators gain tools for audits, with fines up to INR 250 crore for violations. The framework integrates with Digital Public Infrastructure, enabling secure AI in Aadhaar-linked services.
Seven Sutras AI Governance India: Ethical Foundations
The seven sutras AI governance India draw from the Reserve Bank of India’s FREE-AI report. Adapted for broader sectors, these principles guide ethical AI. They blend Indian values of trust and equity with global norms.
- Trust is the Foundation – Build reliable systems with verifiable outputs to earn public confidence.
- People First – Centre human rights, ensuring AI enhances dignity and accessibility.
- Innovation over Restraint – Promote agile development without stifling creativity.
- Fairness and Equity – Eliminate biases to foster inclusive outcomes across demographics.
- Accountability – Assign clear responsibilities across the AI value chain.
- Understandable by Design – Prioritise explainability for users and overseers.
- Safety, Resilience and Sustainability – Design robust models that withstand shocks and minimise environmental impact.
These seven sutras AI governance India form the philosophical core. Prof. B. Ravindran noted the committee adopted them as a “sound basis for building future regulation.” They apply to foundation models like BharatGen, India’s multilingual LLM.
The sutras underpin six pillars: governance structures, capacity building, risk mitigation, accountability, institutions, and innovation enablement. This holistic view addresses the full AI lifecycle, from data sourcing to deployment.
Regional Implications for South Asia
South Asia’s AI ecosystem interconnects tightly. India hosts data centres serving 40% of regional cloud traffic. The India AI governance guidelines will influence exports, requiring compliant models for access. Pakistan’s fintech sector, reliant on Indian APIs, must align to avoid disruptions.
Bangladesh and Nepal seek partnerships under the framework’s reciprocity clause. It offers technical aid for adopting similar standards, potentially via SAARC forums. Sri Lanka’s e-governance initiatives could integrate the risk tools, reducing cross-border cyber threats.
The guidelines advocate Global South leadership. Abhishek Singh, CEO of IndiaAI Mission, said: “These guidelines will be a cornerstone as we move forward in developing AI for India, and they can actually become a role model for AI governance globally.” This extends to joint sandboxes for regional challenges like climate modelling.
Data localisation rules tighten, but adequacy pacts ease flows. Regional startups gain from shared datasets under equitable terms, boosting collective GDP by 15% via AI by 2030, per mission estimates.
Background
The IndiaAI Mission launched in March 2024 with seven pillars: compute, datasets, skills, startups, applications, and safe AI. Governance forms the eighth implicit layer. Early efforts included a March 2024 advisory on untested models, later softened amid feedback.
Subsequent drafts evolved through stakeholder inputs. The RBI’s August 2025 FREE-AI report provided the sutra template, tailored here for non-financial use. Consultations involved NITI Aayog, Microsoft Research, and iSPIRT.
MeitY’s approach avoids a standalone AI law, favouring amendments to existing statutes. This mirrors EU’s risk tiers but emphasises innovation, suiting India’s 1.4 billion users.
What’s Next
Implementation starts with AIGG formation by December 2025. Pilot sandboxes test high-risk AI in health by Q2 2026. The India AI Impact Summit in February 2026 will showcase prototypes.
International tie-ups with EU and GPAI test interoperability. MeitY eyes cabinet nods for funding boosts. Long-term, the framework evolves via annual reviews. These India AI governance guidelines pave the way for equitable AI leadership in South Asia.
Published in SouthAsianDesk, November 6th, 2025
Follow SouthAsianDesk on X, Instagram, and Facebook for insights on business and current affairs from across South Asia.
