IND-008

UIDAI–Sarvam AI Voice Services for Aadhaar: AI-Enabled Multilingual Voice Interactions and Real-Time Fraud Detection

Download PDF
India South Asia Lower middle income Operational Deployment (Limited Rollout) Confirmed

Unique Identification Authority of India (UIDAI), Ministry of Electronics and Information Technology, Government of India

At a Glance

What it does LLMs for content creation, transformation and modality conversion — User communication and interaction
Who runs it Unique Identification Authority of India (UIDAI), Ministry of Electronics and Information Technology, Government of India
Programme Aadhaar (Unique Identification Programme)
Confidence Confirmed
Deployment Status Operational Deployment (Limited Rollout)
Key Risks Not assessed
Source Quality 4 sources — News article / media, Government website / press release

The Unique Identification Authority of India (UIDAI) has partnered with Bengaluru-based Sarvam AI, an indigenous full-stack Generative AI company, to deploy AI-enabled voice-based interaction and real-time fraud detection services for Aadhaar, India's national biometric identity programme covering over 1.3 billion enrolled residents. The partnership agreement came into effect on 18 March 2025 and is initially valid for one year, with the option to extend by an additional year (PIB, 2025a).

Sarvam AI has delivered a custom Generative AI stack that performs three core functions within the Aadhaar ecosystem. First, the system conducts AI-driven voice-based interactions for resident-centric use cases, enabling near real-time feedback collection from Aadhaar number holders regarding their enrolment and update processes. This feedback mechanism captures information on the quality of service provided at Aadhaar enrolment and update centres, including the detection and reporting of overcharging of residents by service providers. Second, the system provides real-time fraud alerts to Aadhaar number holders when the AI identifies suspicious activity during authentication requests. The AI continuously monitors authentication transactions and flags irregularities, reducing the risk of identity fraud and unauthorised access to Aadhaar-linked services. Third, the system delivers these capabilities across ten Indian languages — Hindi, English, Telugu, Tamil, Marathi, Gujarati, Kannada, Odia, Punjabi, and Malayalam — with plans to expand language coverage in subsequent months (PIB, 2025a; Elets eGov, 2025).

The technical architecture of the deployment is notable for its emphasis on data sovereignty and security. Sarvam AI's custom GenAI stack is hosted entirely on-premise within UIDAI's air-gapped infrastructure. No data leaves UIDAI's secure environment at any stage of operation, ensuring full compliance with India's data sovereignty and security protocols. The Ministry of Electronics and Information Technology confirmed that the deployment architecture guarantees that all data processing occurs within UIDAI's controlled perimeter. UIDAI retains full ownership of the deployed solution, with the technology having been developed through collaboration between Sarvam AI experts and UIDAI's Technology Centre in Bengaluru under UIDAI's volunteer policy (PIB, 2025a; Inc42, 2025; Elets eGov, 2025).

Sarvam AI, founded in 2023 by Pratyush Kumar and Vivek Raghavan, is an Indian AI startup that has developed indigenous foundation models optimised for Indic languages. The company raised USD 41 million in Series A funding in December 2023, led by Lightspeed Venture Partners with participation from Peak XV Partners and Khosla Ventures. Sarvam AI's technology stack includes Sarvam-1 (also known as Sarvam 2B), a 2-billion parameter open-source language model pre-trained on 4 trillion tokens with approximately 50 per cent dedicated to Indic language data, achieving 2-4x greater token efficiency compared to existing multilingual models. The company has also developed Bulbul, a text-to-speech model available in 11 Indian languages with 39 distinct speaker voices; Saaras, a speech-to-text model supporting all 22 scheduled languages of India including telephony audio and code-mixed speech; and Shuka 1.0, India's first open-source audio language model. Sarvam AI's Conversations platform handles over 100 million interactions with sub-500-millisecond latency across 11 Indian languages. The company was selected as one of 12 organisations under the IndiaAI Mission's Innovation Centre pillar, receiving financial and compute support amounting to Rs 246.72 crore from the Government of India (Inc42, 2025; PIB, 2025b).

The UIDAI partnership represents a significant deployment of indigenous Indian AI technology within critical national digital infrastructure. UIDAI CEO Bhuvnesh Kumar stated that 'UIDAI is a people centric organization. GenAI is the next technology evolution in UIDAI's journey as a technology pioneer, building on our long-standing commitment to innovation.' Vivek Raghavan, Co-Founder of Sarvam AI, stated that 'This engagement exemplifies the immense potential of AI to drive public good.' The Aadhaar system processes hundreds of crores of authentication transactions annually — over 284 crore authentication transactions were recorded in January 2025 alone, with cumulative authentication transactions exceeding 150 billion (15,011.82 crore) by April 2025 — making the fraud detection and voice interaction capabilities operationally significant at national scale (PIB, 2025a; PIB, 2025c).

Classifications follow the DCI AI Hub Taxonomy. Hover over field labels for definitions.

Social Protection Functions

Implementation/delivery chain
Outreach/communications/sensitisation primaryAccountability mechanisms Registration
SP Pillar (Primary) The social protection branch: social assistance, social insurance, or labour market programmes. Social assistance
Programme Name Aadhaar (Unique Identification Programme)
Programme Type The type of social protection programme, classified under social assistance, social insurance, or labour market programmes. View in glossary Other
System Level Where in the social protection system the AI is applied: policy level, programme design, or implementation/delivery chain. View in glossary Implementation/delivery chain
Programme Description Aadhaar is India's national biometric identity programme administered by the Unique Identification Authority of India (UIDAI) under the Ministry of Electronics and Information Technology. The programme issues a 12-digit unique identity number to residents based on biometric and demographic data, serving as foundational digital infrastructure for identity verification, authentication, and e-KYC across government welfare delivery, financial inclusion, and public services. Over 1.3 billion residents are enrolled.
Implementation Type How the AI output is produced: Classical ML, Deep learning, Foundation model, or Hybrid. Affects validation, compute requirements, and governance profile. View in glossary Foundation model
Lifecycle Stage Current stage in the AI lifecycle, from problem identification through to monitoring, maintenance and decommissioning. View in glossary Integration and Deployment
Model Provenance Origin of the AI model: developed in-house, adapted from open-source, commercial/proprietary, or accessed via third-party API. View in glossary Developed in-house
Compute Environment Where the AI system runs: on-premise, government cloud, commercial cloud, or edge/device. View in glossary On-premise
Sovereignty Quadrant Classification of data and compute sovereignty: I (Sovereign), II (Federated/Hybrid), III (Cloud with safeguards), or IV (Shared Innovation Zone). View in glossary I — Sovereign AI Zone
Data Residency Where the data used by the AI system is stored: domestic, regional, or international. View in glossary Domestic
Data Residency Detail Additional detail on the specific data hosting arrangements and jurisdictions. Custom GenAI stack hosted on-premise within UIDAI's air-gapped infrastructure. No data leaves UIDAI's secure environment at any stage of operation. Ministry of Electronics and IT confirmed full data sovereignty compliance.
Cross-Border Transfer Whether data crosses national borders, and if so, whether documented safeguards are in place. View in glossary None
Decision Criticality The rights impact of the decision the AI supports. High criticality requires HITL oversight; moderate requires HOTL; low may operate HOOTL. View in glossary Moderate
Human Oversight Type Level of human involvement: Human-in-the-Loop (active review), Human-on-the-Loop (monitoring), or Human-out-of-the-Loop (periodic audit). View in glossary HOTL
Development Process Whether the AI system was developed fully in-house, through a mix of in-house and third-party, or fully by an external provider. View in glossary Mix of in-house and third-party
Highest Risk Category The most significant structural risk source identified: data, model, operational, governance, or market/sovereignty risks. View in glossary Not assessed
Risk Assessment Status Whether a formal risk assessment, informal assessment, or independent audit has been conducted for this system. Not assessed

Risk Dimensions

Data-related risks
Market, sovereignty and industry structure risks

Impact Dimensions

Equality, non-discrimination, fairness and inclusion
Systemic and societal
  • Data minimisation controls
CategorySensitivityCross-System LinkageAvailabilityKey Constraints
National ID and biometric databasesSpecial categoryLinks data across multiple systemsCurrently available and usedAadhaar authentication transaction data including biometric and demographic records of 1.3+ billion enrolled residents; used for fraud detection monitoring during authentication requests
Unstructured and text-based contentPersonalSingle source (no linkage)Currently available and usedVoice-based feedback from Aadhaar holders on enrolment and update experiences collected in 10 Indian languages; includes complaints about overcharging by service providers

Elets News Network (2025) 'UIDAI Introduces AI-Powered Voice Interactions and Fraud Detection for Aadhaar Services', Elets eGov, 24 March. Available at: https://egov.eletsonline.com/2025/03/uidai-introduces-ai-powered-voice-interactions-and-fraud-detection-for-aadhaar-services/ (Accessed: 25 March 2026).

View source News article / media

Choudhary, L. (2025) 'UIDAI Partners Sarvam AI To Offer AI Voice, Fraud Detection To Aadhaar', Inc42, 18 March. Available at: https://inc42.com/buzz/uidai-partners-sarvam-ai-to-offer-ai-voice-fraud-detection-to-aadhaar/ (Accessed: 25 March 2026).

View source News article / media

Press Information Bureau (2025a) 'UIDAI partners with indigenous GenAI Company Sarvam AI to enhance user experience of Aadhaar services', Press Release PRID 2112485, 18 March. New Delhi: Government of India.

View source Government website / press release

Press Information Bureau (2025b) 'Sarvam AI Powering a Made-in-India AI Revolution', Press Release PRID 2231169. New Delhi: Government of India.

View source Government website / press release
Deployment Status How far the system has progressed into real-world operational use, from concept/exploration through to scaled and institutionalised. View in glossary Operational Deployment (Limited Rollout)
Year Initiated The year the AI system was first initiated or development began. 2025
Scale / Coverage The scale and geographic or population coverage of the deployment. National — Aadhaar covers 1.3+ billion enrolled residents. System supports voice interaction and fraud detection in 10 Indian languages with planned expansion. Aadhaar processes 284+ crore authentication transactions monthly (January 2025), with cumulative authentications exceeding 150 billion.
Funding Source The source(s) of funding for the AI system development and deployment. UIDAI / Government of India. Sarvam AI contributed development under UIDAI's volunteer policy. Sarvam AI separately receives Rs 246.72 crore in financial and compute support under IndiaAI Mission.
Technical Partners External technology vendors, academic partners, or development partners involved. Sarvam AI (Bengaluru-based indigenous GenAI startup, founded 2023 by Pratyush Kumar and Vivek Raghavan). USD 41 million Series A funding (December 2023) from Lightspeed Venture Partners, Peak XV Partners, and Khosla Ventures. Technology stack includes Sarvam-1/2B (2B-parameter Indic language model), Bulbul (TTS in 11 languages), Saaras (STT for 22 scheduled languages), Shuka 1.0 (audio LM). Selected as one of 12 organisations under IndiaAI Mission Innovation Centre pillar.

How to Cite

DCI AI Hub (2026). 'UIDAI–Sarvam AI Voice Services for Aadhaar: AI-Enabled Multilingual Voice Interactions and Real-Time Fraud Detection', AI Hub AI Tracker, case IND-008. Digital Convergence Initiative. Available at: https://socialprotectionai.org/use-case/IND-008 [Accessed: 1 April 2026].

Change History

Updated 1 Apr 2026, 08:11
by system (system)
Created 30 Mar 2026, 08:40
by v2-import (import)