DPI Deep Dive — Friday | July 24, 2026

Layer L5: Sectoral Infrastructure — ABHA & Digital Health, AgriStack, eCourts

This week’s sectoral DPI layer saw significant movement across health, agriculture, and judiciary. India is hosting the BRICS Health Ministerial with digital health as a centrepiece, NHA conducted a major national review of ABDM, AgriStack-ULI convergence with land records accelerated, and the Supreme Court’s draft AI regulations for courts approached finalisation. Here’s the deep dive.

1. BRICS Health Ministerial: India Positions ABDM as a Global Digital Health Template

India convened the 16th BRICS Health Ministers’ Meeting in Chandigarh from July 22–24, 2026, under its BRICS Chairship. Digital health cooperation was a primary agenda item, with a dedicated high-level side event on July 24 titled “BRICS Dialogue: From Digital Health Records to AI-Enabled Healthcare — Strengthening the Continuum of Care through Interoperable Data Infrastructure and Responsible Artificial Intelligence.” ¹

This is diplomatically significant. India is effectively using BRICS as a platform to export the ABDM model — ABHA IDs, Health Information Exchange, HIP-compliant registries — to other large, diverse economies. The side event explicitly ties interoperable data infrastructure (the ABDM playbook) with responsible AI in healthcare, framing India’s approach as a complete stack other nations can adopt.

The Aarogya Setu 2.0 launch on June 29 — repurposing the COVID-era app into a comprehensive Personal Health Record platform with ABHA creation, Scan & Share, PM-JAY wallet integration, and AI-powered medical record digitisation ² — gives India a concrete citizen-facing product to showcase. With nearly 20 crore existing downloads, the conversion from pandemic contact-tracing app to universal health gateway is one of the most ambitious digital pivots in any country’s DPI stack.

Cross-layer connection: The BRICS digital health dialogue sits at the intersection of L1 (identity — ABHA as health ID), L2 (payments — PM-JAY insurance claims), and L5 (sectoral health infrastructure). It also connects to L7 (trust) since interoperable health data across nations raises sovereignty and data localisation questions.

2. NHA Chintan Shivir: AI, Analytics, and Interoperability Take Centre Stage

The National Health Authority held a two-day Chintan Shivir (National Review Meeting) on July 17–18 at Madhya Pradesh Bhawan, New Delhi, reviewing progress on both AB PM-JAY and ABDM. Union MoS (Health) Prataprao Jadhav inaugurated the session. ³

Key focus areas included financial sustainability of AB PM-JAY, strengthening public hospitals, and — critically for DPI watchers — AI, health analytics, and digital interoperability. Fifteen awards were presented to states and UTs for excellence in ABDM implementation.

Simultaneously, state-level ABDM capacity building continued: Jharkhand’s Rajendra Institute of Medical Sciences (RIMS), Ranchi, conducted training programmes for junior and senior resident doctors on ABDM standards, and Arunachal Pradesh activated District ABDM Coordinators across multiple districts for ABHA ID creation drives. ⁴

The emphasis on AI and analytics at the Chintan Shivir signals that NHA is moving beyond basic ABHA registration (60+ crore IDs created) toward extracting value from the data layer — using health records for population health analytics, fraud detection in PM-JAY claims, and clinical decision support. This is the natural progression for any DPI: identity first, then use-cases, then intelligence.

3. AgriStack-ULI Convergence: Land Records Meet Digital Lending

The Department of Land Resources held a high-level meeting with the RBI Innovation Hub (RBIH) to explore integrating digital land records with the Unified Lending Interface (ULI), specifically for farmers and rural citizens. ⁵

This is a major cross-DPI convergence story. AgriStack already connects its Farmer Registry and Crop Sown Registry to ULI, enabling paperless, sub-30-minute Kisan Credit Card loans by pulling verified data via APIs. The FM’s Budget 2026 speech called AgriStack “one of the next UPI” initiatives. Now, layering in land records — the ultimate collateral asset in Indian agriculture — creates a three-way data fusion: who the farmer is (Farmer ID), what they grow (Crop Sown Registry), and what they own (digital land records).

The practical impact is significant. Currently, farm credit is often disbursed based on self-reported data, leading to misallocation, multiple loans on the same parcel, and exclusion of legitimate borrowers who lack paper titles. The Hindu Business Line reported that Maharashtra used AgriStack-verified crop data to disburse ₹14,000 crore in disaster relief to 89 lakh farmers in just five days during February 2026 — a concrete demonstration of DPI’s speed advantage over manual processes. ⁶

The new RBI Kisan Credit Card Directions, 2026 (issued June 19, effective January 1, 2027) consolidate KCC rules across commercial banks, SFBs, RRBs, and rural co-operative banks, retaining the ₹2 lakh collateral-free limit. This regulatory consolidation creates a uniform playing field for ULI-enabled digital lending to flow through AgriStack pipes.

Cross-layer connection: This story spans L2 (payments/credit through ULI), L3 (data exchange via API-based consent flow), and L5 (AgriStack as sectoral DPI). The land records integration also touches L1 (Aadhaar seeding for farmer identity).

4. Supreme Court AI Regulations: The Draft Nears Finalisation

The Supreme Court’s AI Committee released draft Regulations for Use of Artificial Intelligence in Courts, 2026 on June 3, inviting public comments by June 20. By mid-July, the Karnataka High Court — which has been among the most proactive in engaging with the framework — extended the deadline for stakeholder comments on its own state-level draft, indicating active deliberation is underway. ⁷

The draft regulations are remarkably principled. Key provisions include:

• Absolute ban on risk scoring: No AI may be used for “assessment of flight risk, prediction of recidivism, evaluation of bail eligibility, or determination of the credibility of parties or witnesses” — a direct rejection of the controversial COMPAS-style algorithms used in some Western jurisdictions. ⁸
• Ban on continuous surveillance of judicial officers, advocates, or litigants using AI.
• Mandatory disclosure when AI “materially assists” in case management, document analysis, or judicial administration — parties must be informed.
• AI Committee governance: The Supreme Court AI Committee serves as the apex body, with High Courts constituting their own committees for local implementation.
• Periodic audits of all court AI systems.

This regulatory framework arrives alongside eCourts Phase III (₹7,210 crore outlay), which was formally launched by CJI Surya Kant in April 2026 with AI at its core. The Supreme Court website already features an AI Assistant powered by VANI (Voice-based AI), and Digital Courts 2.1 — an AI tool for paperless court management — is being deployed to judicial officers. ⁹

The tension is clear: the judiciary wants AI’s efficiency gains (faster case management, automated transcription, intelligent case search) but is explicitly drawing red lines around judicial decision-making. Whether an under-resourced court system can enforce these safeguards — particularly the audit requirements and disclosure mandates — remains the open question.

The SAMADHAN SAMAROH 2026 initiative, which culminates in a Special Lok Adalat on August 21–23, adds another dimension: the Supreme Court has identified 45,098 pending matters for possible mediated resolution, using the eCourts infrastructure to manage the process at scale. ¹⁰

5. Aarogya Setu 2.0 Adoption Picks Up: States Ramp Up ABHA Onboarding

Three weeks after the Aarogya Setu 2.0 launch, state-level ABDM adoption is visibly accelerating. The ABDM J&K office promoted ABHA creation via Aarogya Setu 2.0 in multiple awareness campaigns. Andaman & Nicobar Islands’ State Health Agency reported creating 138 ABHA IDs in a single workshop on July 10. ¹¹

The AHA (Association of Healthcare Providers India) noted on July 18 that “Ayushman Bharat Digital Mission standards are being leveraged to reduce cost of healthcare innovation” — suggesting that private-sector healthcare providers are beginning to see ABDM interoperability standards not as compliance burden but as cost-reduction infrastructure. ¹²

With 60+ crore ABHA IDs already created and Aarogya Setu 2.0 providing a mass-market on-ramp, the next inflection point will be actual health record linkage — getting hospitals and diagnostic labs to consistently push records into patients’ ABHA-linked PHRs. The Scan & Share feature (tokenless OPD registration via QR) is the primary hook, but hospital adoption remains uneven.

This Week’s DPI Sectoral Scorecard

1. BRICS Health Ministerial: India Pitches ABDM as a Global Digital Health Template

India convened the 16th BRICS Health Ministers’ Meeting in Chandigarh from July 22-24, 2026, with digital health and AI-enabled healthcare at the top of the agenda. ¹

On July 24, India hosted a dedicated side event titled “BRICS Dialogue: From Digital Health Records to AI-Enabled Healthcare — Strengthening the Continuum of Care through Interoperable Data Infrastructure and Responsible Artificial Intelligence.” The framing is telling: India isn’t just participating in BRICS health cooperation — it’s positioning ABDM (Ayushman Bharat Digital Mission) as an exportable DPI model for the Global South.

The ABDM stack — ABHA health IDs, Health Information Exchange, personal health records via Aarogya Setu 2.0 — is being presented to BRICS nations as a replicable blueprint. Over 60 crore ABHA IDs have been created, and with Aarogya Setu 2.0’s relaunch on June 29 as a full PHR (Personal Health Record) platform built on its 20 crore user base, India now has a consumer-facing gateway to match the backend infrastructure.

Cross-layer connection: This is L5 (Health DPI) connecting to L2 (Payments — PM-JAY insurance claims flow through UPI/RuPay rails) and L3 (Data Exchange — ABDM’s Health Information Exchange uses consent-based data sharing protocols analogous to DigiLocker’s architecture). When India pitches ABDM to BRICS, it’s implicitly pitching the entire seven-layer DPI stack.

What to watch: Whether any BRICS nation signs a formal MoU to adopt ABDM-adjacent standards. A successful export would validate the “DPI as diplomatic instrument” thesis that India has been building since UPI went international.

2. NHA Chintan Shivir: ABDM’s Mid-Life Review and the Push Toward AI

The National Health Authority conducted a two-day Chintan Shivir (National Review Meeting) on July 17-18 in New Delhi, reviewing both AB PM-JAY (insurance) and ABDM (digital health infrastructure). ²

Fifteen awards were presented to states and UTs for excellence in ABDM implementation. Key focus areas discussed:

• Financial sustainability of AB PM-JAY — the insurance scheme that covers 50 crore+ beneficiaries is facing cost pressures as private hospital empanelment grows.
• AI integration in health analytics — using ABDM’s data layers for predictive health interventions, fraud detection in insurance claims, and population health management.
• Digital interoperability between ABDM and state-level health systems, which remain patchy despite the central stack being mature.
• Strengthening public hospitals through digital tools — connecting government facilities to the ABDM network has been slower than private hospital onboarding.

Simultaneously, state-level ABDM implementation is accelerating. Jharkhand’s Rajendra Institute of Medical Sciences conducted a capacity-building training programme for doctors on ABDM standards, and Arunachal Pradesh’s district ABDM coordinators met to review progress. The Aarogya Setu 2.0 ABHA onboarding push in J&K also continued with “Skip the Queue” campaigns at government hospitals.

Cross-layer connection: ABDM’s AI push connects directly to L7 (Security & Privacy) — as health data becomes the fuel for AI analytics, the DPDP Act’s enforcement framework (still being operationalised) will determine what’s permissible with citizen health data. The NHA’s focus on “digital interoperability” also touches L3 (Data Exchange) — ABDM’s APIs must talk to state health IT systems that were built on entirely different stacks.

3. AgriStack-ULI Convergence: Land Records Integration for Sub-30-Minute Farm Loans

The Department of Land Resources held a high-level meeting with the RBI Innovation Hub (RBIH) this week to explore integrating digital land records with the Unified Lending Interface (ULI). ³ This is a significant convergence moment between L5 (AgriStack) and L2 (Payments/Financial Rails).

The premise is straightforward but powerful: AgriStack’s Farmer Registry and Crop Sown Registry currently provide verified farmer identity and sowing data. ULI, built by RBIH, pulls together consensual data from multiple DPI layers to enable instant lending decisions. Adding digital land records to this pipeline would give lenders verified ownership data — eliminating the single biggest collateral verification bottleneck in agricultural lending.

Current AgriStack scale: ~8.62 crore Farmer IDs created (against an 11 crore target), Digital Crop Survey covering 25+ crore plots across 492 districts, and geo-referenced village maps covering 5.4 lakh of 6.75 lakh villages. The RBI Kisan Credit Card Directions, 2026 (issued June 19, effective January 1, 2027) consolidate KCC rules and retain the ₹2 lakh collateral-free limit — but land-record-backed ULI loans could go well beyond that ceiling with lower risk.

Maharashtra’s February 2026 disaster relief disbursement — ₹14,000 crore to 89 lakh farmers in five days using AgriStack data — remains the proof-of-concept for what this infrastructure enables at scale. ⁴

Cross-layer connection: This is a three-layer convergence: L5 (AgriStack) + L2 (ULI/credit rails) + L1 (Aadhaar-seeded land records for identity verification). The DoLR’s own emphasis on “Aadhaar seeding for accurate farmer identification” signals that L1 identity infrastructure remains the foundational trust layer for every sectoral DPI.

4. Supreme Court AI Regulations: The Comment Period Closes, What Comes Next

The Supreme Court’s draft “Regulations for Use of Artificial Intelligence in Courts, 2026,” released on June 3, had a public comment deadline that generated significant debate through July. ⁵ The Karnataka High Court extended its own stakeholder consultation period, with its latest circular (dated July 9, 2026) noting continued public engagement. ⁶

The draft regulations are remarkably prescriptive for a first-of-its-kind framework:

• Absolute ban on risk scoring: No AI may assess flight risk, recidivism probability, bail eligibility, or witness credibility. This directly addresses the COMPAS-style predictive policing failures seen in the US.
• Ban on continuous surveillance: AI cannot monitor judges, advocates, or court staff.
• Mandatory disclosure: Courts must inform litigants when AI tools “materially assist” in case management, document analysis, or judicial administration.
• AI Committee governance: A Supreme Court AI Committee serves as the apex body, with High Courts forming their own committees for local implementation.
• Periodic audits: All court AI systems must undergo regular auditing, though enforcement mechanisms in an under-resourced judiciary remain an open question.

Why this matters for DPI: eCourts Phase III (₹7,210 crore outlay, launched April 2026) is explicitly AI-forward — with Digital Courts 2.1, AI-assisted case file reading, and the VANI AI assistant already piloted on the Supreme Court website. The AI regulations aren’t abstract governance theatre; they’re the guardrails for infrastructure that’s already being deployed.

The SAMADHAN SAMAROH 2026 — the Supreme Court’s nationwide mediation initiative running from April to August — is another eCourts Phase III programme building on digital infrastructure. District courts across India are actively enrolling pending cases for the Special Lok Adalat scheduled for August 21-23. The Supreme Court has identified 45,098 pending matters across multiple categories for possible amicable resolution. ⁷

Cross-layer connection: Court AI governance (L5 eCourts) intersects with L7 (DPDP Act — what training data is used for judicial AI?), L3 (interoperability — can AI tools access eCourts APIs?), and L6 (governance — the AI Committee structure mirrors DARPG’s e-governance frameworks).

5. Aarogya Setu 2.0: From Pandemic Tracing App to Universal Health Gateway

While Aarogya Setu 2.0 was launched on June 29 (just outside the 7-day window), its adoption momentum has been the dominant ABDM story this week. ⁸ The app’s transformation from a COVID-19 contact-tracing tool into a comprehensive Personal Health Record platform represents one of the most ambitious repurposing exercises in global digital governance.

Key capabilities now live:

• ABHA creation and management directly within the app
• AI-powered digitisation of medical records with personalised health insights
• PM-JAY Wallet integration for insurance claim tracking
• Family health profile management under a single account
• Scan & Share for paperless healthcare interactions (using ABDM’s interoperable standards)
• Integration with e-RaktKosh for real-time blood availability
• Wearable device integration for health tracking

The strategic rationale is clear: ABDM has built mature backend infrastructure (Health ID, HIE, HIP compliance), but citizen adoption lags. Aarogya Setu’s 20 crore existing installs provide an immediate distribution channel that bypasses the cold-start problem. It’s the same playbook as UPI riding on existing mobile banking apps — embed the new DPI layer inside something people already use.

The consumer angle: For citizens, the question is whether a single app can be both trusted (Aarogya Setu carries pandemic-era surveillance baggage) and useful enough to change behaviour. The health record digitisation feature — snapping photos of prescriptions and lab reports — is the stickiest use case. If it works well, ABHA adoption could accelerate dramatically.

Cross-Layer Analysis

This week’s L5 developments reveal three structural patterns in India’s sectoral DPI:

1. Export ambition: BRICS is the staging ground for taking ABDM global. India is no longer just building DPI for domestic use — it’s building reference implementations for diplomatic leverage.

2. Convergence is accelerating: AgriStack-ULI-land records is a three-layer DPI convergence that would have been unthinkable two years ago. The pieces (Aadhaar, land digitisation, ULI, AgriStack registries) existed separately; now they’re being wired together.

3. Governance is racing to catch up: The Supreme Court’s AI regulations for eCourts show that governance frameworks are being drafted in parallel with deployment, not after. Whether they can be enforced is the open question — but the intent to regulate is genuine.

The sectoral layer is where DPI touches lives most directly — a farmer getting a loan in 30 minutes, a patient accessing their full health history on their phone, a litigant resolving a dispute without years of waiting. This week showed all three use cases advancing, albeit at different speeds.