This article is authored by Kalyan Kumar, President, HCLSoftware; Amitabh Nag, Chief Executive Officer, Digital India BHASHINI Division; Leon Martinez, Thematic Lead, GovTech, World Economic Forum; and Achyut Chandra, GovTech and Open Innovation Lead, HCLSoftware, HCLTech.

As governments embrace AI to transform public services, a bigger question is emerging: can technology make citizens feel more heard, understood and included? From grievance redressal to multilingual access, India’s experience offers a glimpse of how AI, thoughtfully governed, can help build public services around trust, responsiveness and the realities of everyday citizens.

AI can help governments become more accessible, inclusive and trusted, but for that it needs to be designed around citizens’ realities, its adoption hand-in-hand with strong governance, human judgement and institutional accountability. In multilingual societies, one of the most basic tests of public value is whether citizens can understand, access and use government services in the language they speak.

Public good is ultimately measured by the value citizens experience receiving in their everyday interactions with government: whether services are accessible, grievances are addressed fairly, and institutions remain accountable for outcomes. Citizen trust growth depends on governments being responsive, transparent and able to act on people’s needs at scale. Digital governance provides the systems, data flows and institutional coordination required to make this possible, while public service delivery is where citizens experience its impact most directly.

The main argument of the World Economic Forum’s GovTech Compass is that government technology should be guided by public value, trust and responsible implementation, not efficiency alone. In this context, AI-based tech solutions can help governments listen better, identify recurring service failures, route requests faster and convert citizen feedback into actionable insights.

India's grievance redressal journey offers a case in point, showing how AI-enabled public services can underpin more responsive grievance redressal while reinforcing a central lesson for governments everywhere: the goal is not automation for its own sake, but better and fairer citizen outcomes. Similarly, India's BHASHINI demonstrates how AI can serve as public-good infrastructure by reducing language barriers and making digital services more accessible to citizens in the languages they use every day. For emerging countries, the opportunity lies in adapting these lessons to build trust-centric digital governance that is inclusive, accountable and capable of responding at population scale.

The Paradox of Progress

In much of the world, citizens still struggle to make their voices heard, even as governments invest heavily in digital transformation and emerging technologies.

The UN E-Government Survey 2024 findings highlight that while online services have spread, few governments have fully integrated citizen feedback into accountability processes. Meanwhile, AI is already demonstrating what is possible: routing complaints in seconds, detecting recurring patterns that signal systemic failures, and analysing feedback that would be difficult for human teams to manage alone. The gap between technological possibility and administrative reality has never been more important to close.

If designed toward the public good, AI can bridge this divide — helping institutions listen, respond, and act at scale. But technology alone is not enough. The governance decisions built into architecture around it — who controls data, how decisions are explained, whether citizens have meaningful recourse — matter just as much.

What Constitutes Public Good and How Can Technology Aid It?

Public good is ultimately reflected in the value citizens experience through their interactions with the government. The World Economic Forum’s GovTech Compass argues that public value is created when technology is designed around citizens’ lived realities and shaped through inclusive decision-making across the full GovTech and digital public infrastructure life cycle, ensuring that solutions respond to real needs, strengthen accountability and build trust. In practice, this means delivering services that are responsive, transparent and accountable to citizen needs. As populations grow and expectations rise, governments face increasing pressure to deliver these outcomes at scale. Technology therefore becomes more than a tool for digitisation-it becomes an enabler of better public service delivery, stronger institutions and more trusted governance.

What AI Can Do

Successful GovTech deployment requires a structured, phased implementation approach. Governments must first establish digital foundations, then create interoperable systems and trusted data flows, before realising the full potential of AI and more advanced autonomous capabilities. The GovTech Progression Matrix below maps this evolution across three stages — Digitisation, Digitalisation and Agentic — showing how technology, data, governance and AI maturity develop together to create greater public value and citizen trust.

Figure: GovTech Progression Matrix 

At the digitisation stage, AI can support low-level but high-volume tasks such as scanning petitions, transcribing calls, translating submissions, extracting metadata, removing duplicate records and creating searchable grievance histories. These interventions reduce manual effort and help governments build the structured data foundation required for future AI use.

At the digitalisation stage, AI becomes more operationally valuable. Natural language processing can classify complaints and route them to the correct department, dashboards can surface recurring service failures, sentiment analysis can distinguish formal closure from genuine citizen satisfaction, and predictive analytics can anticipate complaint spikes linked to seasonal outages, payment delays or service bottlenecks. This second stage is where AI begins to improve speed, transparency and accountability.

At the agentic stage, AI can move from task support to bounded workflow orchestration. Systems may help coordinate actions across departments, recommend escalation pathways, draft citizen responses, trigger field inspections or prioritize urgent cases. But this stage requires mature data systems, stable APIs, clear accountability, explainability, rollback mechanisms and human oversight. The lesson is simple: governments should not jump to autonomy before building the foundations of trust.

Across these platforms, AI capabilities stand out as highest-value for grievance systems:

• Routing and classification: Natural Language Processing (NLP) models can route complaints to the correct department in seconds, reducing misdirection.
• Systemic pattern detection: When identical complaints arrive from across districts, it signals a policy gap or resource failure. AI can surface these patterns across millions of cases -enabling proactive governance, not just reactive case management.
• Sentiment analysis: AI-assisted analysis of feedback and call transcripts can distinguish complaints that are formally closed from those genuinely resolved — the insight that drove India’s AI+HI pivot. India’s "AI + HI" (Artificial Intelligence + Human Intelligence) pivot represents a strategic national transition from being a back-office IT services provider to a global "brain office" and architecture leader, 
• Predictive intervention: Grievance data analysed over time can anticipate complaint spikes — seasonal outages, payment delays — enabling pre-emptive action.
• Multilingual access: Platforms such as BHASHINI can help citizens submit grievances, access public information and understand government responses in their preferred language, reducing exclusion caused by language barriers.
• Voice-first service delivery: BHASHINI’s speech-to-text and text-to-speech capabilities can support citizens who are less comfortable with written digital interfaces, enabling them to ask questions, receive guidance and interact with services through voice.

Yet these AI capabilities can create public value only if citizens are able to engage with the government in the first place. In multilingual societies, language becomes the foundational interface through which AI-enabled public services are accessed. This is where India's BHASHINI plays a transformative role.

Introducing BHASHINI as India's Language Digital Public Infrastructure

India's BHASHINI (BHASHa INterface for India), developed by the Ministry of Electronics and Information Technology (MeitY), establishes language as a foundational layer of Digital Public Infrastructure (DPI). This AI-powered platform enables every citizen to access digital public services in the language they understand, speak, and trust.

Built on the principle that AI should adapt to India's linguistic diversity, BHASHINI enables multilingual, voice-first public services, ensuring citizens, regardless of language, literacy or geography, can participate in the digital economy.

As India's Language DPI, BHASHINI provides a full-stack Voice AI Suite and multilingual AI services that enable end-to-end voice, text, and document processing across government services. This open capability stack allows governments and ecosystem partners to build scalable, multilingual digital services while ensuring citizens can access public services in the language they understand best.

Supporting 36+ Indian text languages, 23 Indian voice languages and 35 international languages, BHASHINI is an open, interoperable Language DPI ecosystem that enables governments and partners to build multilingual digital solutions. Today it powers 800+ government websites, supports 359+ AI language models, has delivered 8+ billion cumulative AI inferences, and processes 20+ million AI inferences daily.

Its underlying National Hub for Language Technology (NHLT) provides secure, scalable and low-latency infrastructure for real-time multilingual AI services at national scale.

Beyond translation, BHASHINI enables voice-first governance, allowing citizens to access welfare, healthcare, education, justice and grievance redressal services simply by speaking in their preferred language. This strengthens last-mile service delivery, particularly for rural communities, elderly citizens and first-time digital users.

By embedding language as a foundational layer of Digital Public Infrastructure, rather than treating it as a translation feature, BHASHINI demonstrates how multilingual AI can enable inclusive digital governance, strengthen citizen trust, and ensure equitable access to public services. Its open, interoperable Language DPI model provides a blueprint for multilingual societies worldwide.

Scaling with Trust: Guardrails in the Global South

For governments across the Global South, India's grievance redressal journey — and BHASHINI's model of language-inclusive public infrastructure — offers more than a case study; it offers a practical playbook for scaling AI-enabled public service delivery responsibly. The lesson is not to replicate these systems mechanically, but to sequence the transformation carefully: digitize citizen feedback, structure grievance data, connect departments through interoperable workflows, introduce AI for routing and pattern detection, and retain human judgment at critical decision points. This progression matters because AI creates public value only when it is built on reliable data, institutional coordination and clear accountability.

The World Bank's GovTech Maturity Index underscores the importance of modular, interoperable architectures in helping governments link citizen feedback with policy execution. For countries still building digital foundations, the priority should be to create trusted channels for citizens to be heard, make grievance data usable, and establish transparent workflows before moving toward advanced AI capabilities. India's grievance redressal experience demonstrates how governments can move from complaint processing to feedback intelligence — using citizen inputs not only to resolve individual cases but to identify systemic service failures and improve policy implementation. BHASHINI adds a further dimension: ensuring that this intelligence is gathered from all citizens, not just those fluent in dominant languages.

Crucially, infrastructure alone does not create trust. A Global South playbook for AI-enabled grievance redressal must include clear service standards, explainable AI-assisted routing, human review, audit trails, appeal mechanisms and regular checks for bias or exclusion. Citizens ought to know how their complaint moves through the system, who is responsible for resolving it and how to escalate if it stalls. Furthermore, digital-first must never stand for digital-only: the treatment of citizens without reliable internet access cannot be treated as an afterthought.

Moving Towards Trust-Centric Digital Governance

The goal of introducing AI into public services is not to replace the relationship between citizens and the state, but to strengthen it. When grievances are addressed promptly, services are delivered transparently, and governments remain accountable for outcomes, trust is built one interaction at a time. India's experience — from grievance redressal systems that turn complaints into policy intelligence, to BHASHINI's mission of making every government service speakable in every citizen's mother tongue — demonstrates how technology, combined with human judgment and institutional accountability, can help governments listen better, respond faster, and serve citizens more equitably.

As governments worldwide move from digitization to intelligent service delivery, the lesson from India is clear: The future of AI in government will not be judged by the sophistication of its algorithms, but by whether every citizen-regardless of language, literacy, or geography-can be heard, understood, and served.