India today stands at a decisive moment in its governance journey. Artificial Intelligence is no longer a distant technological promise; it is becoming an essential instrument of public administration. With the launch of the IndiaAI Mission, we have set out to ensure that this transformation is purposeful, inclusive and anchored in democratic accountability.
The IndiaAI Mission has been conceived as a whole-of-government and whole-of-society effort to realise the vision of “AI in India” and “AI for India and the world”. Anchored by the Ministry of Electronics and Information Technology (MeitY), the mission integrates compute infrastructure, datasets, skills development, startup enablement and governance frameworks into a single national platform. Over the next five years, this initiative is expected to fundamentally shift governance from document-driven, file-based processes to data-driven, anticipatory and language-inclusive public services delivered at population scale.
THREE OBJECTIVES
At the heart of this mission lie three interconnected objectives. First, we are committed to ensuring that the benefits of artificial intelligence reach every citizen, strengthening public service delivery and promoting inclusive growth. Second, we see AI as a powerful engine of economic growth, capable of improving productivity, enabling the creation of new industries and generating high-skilled employment. Third, we are focused on building sovereign AI capabilities—from large foundational models to domain-specific applications—so that India emerges as a credible and competitive player in the global AI landscape.
THREE PHASES IN THREE YEARS
We have deliberately structured the IndiaAI Mission in phases to ensure early impact as well as long-term institutional strength. Phase One, covering 2025–26, focuses on mission orientation with clearly defined goals, timelines and measurable outcomes. Phase Two, spanning 2026–27, is centred on institutional setup and governance design. Even within this initial three-year window, tangible changes are already visible.
The IndiaAI Compute Portal is now operational, providing access to over 38,231 GPUs at highly subsidised rates, averaging ₹65 per hour. We have released AI Governance Guidelines that place the principle of “Do No Harm” at the centre of all government AI deployments. At the same time, multilingual digital services are being rolled out through BHASHINI across all 22 scheduled Indian languages, significantly reducing language barriers in accessing government services. Alongside this, the safe and trusted AI pillar is operationalising bias detection mechanisms, algorithmic audits and privacy-enhancing technologies to ensure responsible AI deployment from the outset.
AI IN HEALTH EDUCATION & AGRICULTURE
Several sectors have already demonstrated the transformative potential of AI in governance. In healthcare, platforms such as Aarogya Setu and Co-WIN illustrated how AI-enabled systems can operate at unprecedented scale, supporting public health surveillance and vaccine delivery during the pandemic.
In agriculture, AI-driven advisory services have been successfully piloted across states such as Himachal Pradesh, Uttarakhand, Rajasthan, Gujarat and Maharashtra, spanning diverse agro-climatic zones from the Himalayas to arid and plains regions. These solutions assist farmers with crop advisories, pest surveillance and yield prediction tailored to local conditions.
The education sector has also shown remarkable innovation through partnerships involving AICTE, UGC and BHASHINI’s Anuvaadini application, which is translating undergraduate and postgraduate textbooks—including technical content—into multiple Indian languages and making them widely accessible through platforms such as e-KUMBH.
AI IN CBDT & CBIC
What unites these successful deployments is not technological novelty alone, but a consistent focus on solving real problems at scale. They leverage India’s existing digital public infrastructure, prioritise multilingual accessibility and remain firmly oriented towards citizen outcomes rather than technology for its own sake. Advanced applications of AI are also visible within key departments of government. Tax administrations under the Ministry of Finance, including CBDT and CBIC, use sophisticated AI and machine learning models for risk-based analysis of returns, detection of fraudulent GST claims and identification of potential tax evasion.
In the Ministry of Agriculture and Farmers’ Welfare, AI is being deployed through ICAR and startup partnerships for soil health monitoring, crop-yield prediction, pest surveillance and farm-specific advisories. Digital Public Infrastructure platforms under MeitY—such as DigiLocker, UMANG, BHASHINI and IndiaAI-linked pilots—are embedding AI for document classification, conversational access and real-time language translation.
FUTURE PLANS
Looking ahead, I see immense untapped potential in expanding voice-based AI interfaces to bridge both language and digital divides. Conversational AI can enable non-literate citizens to access government services in their own dialects, extending well beyond current language coverage.
Predictive governance represents another major opportunity, where AI can support data-driven decision-making in disaster management, resource allocation and proactive service delivery, often before citizens articulate their needs. The expansion of AI infrastructure into Tier 2 and Tier 3 cities will further catalyse the development of localised AI solutions, including region-specific crop advisories and telemedicine platforms tailored to local health challenges.
MAKING LIFE EASIER
Artificial intelligence is already simplifying everyday government interactions, especially for non-digital and low-literacy users. AI-powered translation, voice interfaces and document understanding allow citizens to engage with public services through speech and local languages rather than complex forms. Voice-to-voice models such as BHASHINI are transforming access by providing multilingual digital services in all scheduled languages using advanced natural language processing, with voice emerging as the primary medium to overcome literacy barriers.
Platforms such as the multilingual e-Gram Swaraj, launched in 2024, ensure that citizens can access digital services in their own language, while BHASHINI’s deployment at the Maha Kumbh 2025 demonstrated how AI can bridge linguistic divides for millions, including rural visitors. For non-digital users, voice-based AI eliminates the need for typing or navigating complicated portals, enabling simple, spoken interactions with government systems.
IN ALL INDIAN LANGUAGES
Language inclusion remains a strategic priority. Mission BHASHINI and BharatGen AI form the backbone of our language strategy. BHASHINI already supports around 20 Indian languages, hosts over 350 AI models and has crossed one million downloads, while BharatGen has been launched as a multimodal large language model supporting all 22 Indian languages. We are placing special emphasis on dialects and underrepresented languages through AIKosh, which hosts language datasets and models, and through partnerships with large public platforms such as the railways to deploy multilingual services at scale.
RISK MITIGATION
As AI becomes embedded in governance, managing risks is non-negotiable. The Committee on India AI Governance Guidelines has recommended a common risk framework emphasising fairness, explainability, human-in-the-loop oversight and sector-specific guardrails for high-risk use cases.
We are aligning AI deployments with the Digital Personal Data Protection (DPDP) Act and the IT Act, mandating privacy-by-design, data minimisation, security audits and impact assessments for sensitive applications. In October 2025, MeitY also introduced comprehensive amendments to the IT Rules, creating India’s first explicit statutory framework for synthetically generated content.
COMPUTE CAPACITY
Compute and data are foundational to India’s AI ambitions. The IndiaAI Compute Portal, operational since March 2025, has rapidly scaled to over 38,000 GPUs through partnerships with multiple cloud service providers. Users can access a wide range of advanced AI chips at highly subsidised rates. We have also launched additional empanelment rounds and issued an RFP to establish public cloud infrastructure, creating a nationwide, distributed AI backbone that democratises access to technology.
Parallelly, the datasets pillar of the IndiaAI Mission focuses on improving data quality and secure access, while ensuring responsible AI principles across the data lifecycle. AIKosh is being developed as a non-personal data and model repository with strong anonymisation, provenance tracking and tiered access policies.
PUBLIC PRIVATE PARTNERSHIP
Startups and industry are central partners in this journey, not vendors at the margins. They are building domain-specific models and last-mile applications, co-creating standards and safety tools, contributing to national datasets and participating in centres of excellence and pilots.
They are also key stakeholders in the emerging AI Safety Institute network, which brings together academia, industry, startups and government to address AI safety challenges in the Indian context. Through the IndiaAI Startup Financing pillar, we are de-risking early-stage deep-tech innovation and crowding in private investment.
CAPACITY BUILDING
Capacity building remains a core focus. We are supporting states and districts through vernacular training programmes, Data and AI Labs in Tier 2 and Tier 3 cities, structured competency frameworks for officials and initiatives such as Responsible AI for Youth. Free, accessible AI courses with Indian examples are being offered through platforms like FutureSkills Prime, while structured training is equipping public officials to apply AI responsibly in policymaking.
India is consciously adopting a principle-based, layered approach to AI governance rather than rushing into standalone legislation. The AI Governance Guidelines operationalise responsible AI through institutional design, voluntary measures and techno-legal instruments, focusing on governing applications rather than technology itself. Accountability remains firmly human. AI systems serve as decision-support tools, while responsibility rests with the officials and institutions that deploy them.
Artificial intelligence is not replacing governance; it is reshaping it. New roles in AI ethics, model auditing, synthetic data generation and human–AI collaboration are emerging within government. Our task is to ensure that this transformation strengthens the state’s capacity to serve citizens with empathy, efficiency and trust. That is the promise of India’s AI journey in governance—and one we are determined to fulfil.
(The author is Secretary, Ministry of Electronics and Information Technology (MeitY) with Government of India. He is 1989-batch IAS officer of Tamil Nadu cadre)
