AI for the Global South: From Governance to Inclusion

Abstract

The panel examined how the Global South can shape an inclusive AI agenda that serves development priorities rather than purely commercial or risk‑avoidance goals. Indian representatives outlined a three‑pillar approach that favours outcome‑oriented, “welfare‑first” AI governance, soft‑touch regulation and broad‑based democratization of AI benefits. Brazilian and African Union delegates stressed the need for south‑south cooperation, digital sovereignty, and multilateral mechanisms such as the UN Global Digital Compact. The discussion moved to concrete Indian infrastructure—the Digital Public Infrastructure (DPI) stack (Aadhaar, Jan‑Dhan, DigiLocker, UPI, etc.)—and how AI can leverage these data assets for agriculture, credit, health and other “last‑mile” services. A short audience Q&A explored challenges of AI literacy, trust, data sovereignty and financing for small‑holder farmers. The session closed with a collective call for coordinated, inclusive governance frameworks that empower people, protect sovereignty and unlock the development potential of AI across the Global South.

Detailed Summary

Dr. Sasmit Patra (India) introduced three “key pillars” that differentiate the Indian (and by extension Global‑South) approach to AI:

Avoidance of iron‑clad, EU‑style legislation – Instead of heavy‑handed rules, India prefers flexible, adaptive governance.
Soft‑touch regulation – Guardrails for safety and security that do not inhibit innovation or data access for researchers.
Welfare‑outcome‑based focus – AI is evaluated by its ability to improve lives at the bottom of the pyramid (farmers, expectant mothers, gig workers, students).

Patra stressed that AI must be democratized, not concentrated in the hands of a few. He gave concrete sectoral examples (precision agriculture, prenatal care, classroom innovation, gig‑worker welfare) before pausing for a brief thank‑you.

Ambassador Lavina Ramkissoon (AU) later echoed the need for an “inclusive global AI agenda” anchored in developmental priorities, but her remarks are largely interwoven with the subsequent discussion and are not transcribed verbatim.

2. India’s Unique Position & Policy Outlook

S. Niranjan Reddy (India) highlighted India’s internal north–south divide, arguing that the country can serve as a “laboratory” for AI applications that benefit its “local south”. He noted:

India has chosen a learning‑by‑observing stance, adapting regulations from other jurisdictions rather than imposing a blanket framework.
The government is reactive, not overly proactive, allowing innovation to flourish while still safeguarding public welfare.
Limited resources but a massive user base require smart, scalable solutions that can be replicated elsewhere.

Reddy positioned India’s model as a potential pilot for other Global‑South nations.

3. Brazil’s Perspective – South‑South Cooperation & Digital Sovereignty

Ambassador Eugênio Vargas Garcia (Brazil) opened with a “South‑Star” metaphor, urging a shift in mindset to view AI through the lens of the Southern Hemisphere. Key points:

AI should be development‑oriented; “development” must appear explicitly in AI governance principles.
The “three sutras” of the AI Impact Summit – People, Planet, Progress – stress inclusion, sustainability, and economic advancement.
Brazil, having chaired BRICS in 2025, has created an AI Task Force and adopted a BRICS‑level statement on AI governance emphasizing digital sovereignty and respect for national policy choices.
A forthcoming digital partnership with India will combine India’s DPI experience with Brazil’s financial‑inclusion tools (e.g., the PICS system). This partnership aims to co‑develop AI‑enabled digital stacks covering data governance, e‑government, semiconductors, and high‑performance computing.

Garcia’s remarks framed AI as a joint South‑South effort that can leverage each country’s comparative advantage while safeguarding sovereignty.

4. The UN Global Digital Compact – Minimum Standards for Inclusive AI

Dr. Sene (UN) explained the Global Digital Compact, a UN‑endorsed framework that establishes a minimum set of standards for digital cooperation, including AI. Highlights:

The Compact is the first global agreement on digital cooperation across all UN member states.
It calls for human‑rights‑based, development‑oriented, peace‑focused baseline standards rather than an exhaustive, restrictive regime.
The UN plans global dialogues (Geneva 2024, New York 2025) and capacity‑building programmes to help member states design national AI strategies that respect these baseline standards.
Sovereignty issues should be addressed layer‑by‑layer, defining where data control starts and ends in each jurisdiction.

Dr. Sene argued that the Compact should prevent fragmentation while allowing flexibility for varied development contexts.

5. Deep‑Dive into India’s Digital Public Infrastructure (DPI)

Dr. Sasmit Patra (again) and Reddy provided a practical walk‑through of the DPI stack:

Stack Component	Function / Example
Aadhaar	Unique biometric ID for every citizen
Jan‑Dhan	Financial inclusion – bank accounts for the un‑banked (≈80 % coverage)
DigiLocker	Digital storage of documents, instantly shareable
UPI	Real‑time payments (≈8 bn transactions/day)
Other Layers	Mudra loans, health portals, agricultural subsidy portals (e.g., PM Kisan Yojana)

Patra illustrated a use‑case: a farmer’s subsidy data (linked to Aadhaar) can be combined with AI to detect leakages, predict pest outbreaks, or flag climate risks, enabling proactive governance. The same stack could generate credit histories for informal vendors via AI‑driven analysis of UPI transaction patterns, then push vernacular loan eligibility notifications through voice bots (leveraging the Bhashini language platform).

Reddy emphasized that state governments are already applying AI at hyper‑local levels (district‑wise cropping patterns, climate alerts) even when operating outside the formal DPI.

6. Audience Q&A – Practical Challenges

A series of audience questions (facilitated by Trisha Ray, moderator) explored three recurring themes:

a. AI Literacy & Trust at the “Last Mile”

Question (Sanhit, policy‑governance researcher) – How can farmers or gig workers trust AI recommendations without technical knowledge?
Answer (Patra) – Emphasised service‑delivery first: AI should deliver tangible benefits (e.g., early cyclone warnings) before demanding user literacy. Once benefits are proven, trust builds organically. Capacity‑building can follow in later phases.

b. Data Sovereignty & Cross‑Border Model Hosting

Question (Reddy) – With generative AI models hosted abroad, how can sovereign data be protected?
Answer (Reddy) – Distinguish government‑relevant data (e.g., policy documents, citizen records) from ordinary user‑generated content. Propose data‑localisation mandates for sensitive datasets while acknowledging the fuzzy legal terrain for “general” data (photos, social‑media posts).

c. Credit Access for Small‑holder Farmers

Question (Nikhil Peter Fernandez, renewable‑energy/agrifood practitioner) – How can AI help create credit scores for farmers lacking formal financial histories?
Answer (Patra & Reddy) – Leverage UPI transaction footprints and DPI‑linked subsidy data to generate an AI‑inferred credit profile, then deliver vernacular loan eligibility messages via bots. The initiative is being explored but not yet formalised; private‑sector pilots may precede a government roll‑out.

d. Role of UN & Non‑State Actors

Question (Fernandez – “non‑state actors” clarification) – Concern about misuse of AI data by malicious actors.
Answer (Dr. Mehdi, UN) – The UN’s role is capacity‑building and awareness, not imposing governance. Member states must decide sovereign policies; the UN can facilitate knowledge‑sharing and global dialogues.

7. Closing Reflections

Each panelist was invited to give a 15‑second final thought:

Dr. Patra – “AI for the Global South is real; we must engage together.”
Reddy – “Multi‑lateral coalitions are essential; open‑source access can bridge resource gaps.”
Garcia – “Place inclusion at the core of governance mechanisms.”
Ramkissoon – “Co‑operation, not competition, will drive equitable AI.”
Ray (Moderator) – Summarised that the conversation would continue over the next three days.

The session concluded with a reminder that the AI Impact Summit will proceed for three more days, building on the themes discussed.

Key Takeaways

Three‑pillar Indian model: soft‑touch regulation, outcome‑oriented welfare focus, and democratization of AI benefits.
Brazil‑India digital partnership will combine India’s DPI expertise with Brazil’s financial‑inclusion tools (PICS) to create a joint AI‑enabled digital stack.
The UN Global Digital Compact sets minimum, human‑rights‑based standards for AI, allowing national flexibility while preventing a fragmented regulatory landscape.
DPI stack components (Aadhaar, Jan‑Dhan, DigiLocker, UPI) provide a rich data foundation; AI can turn these into predictive tools for agriculture, health, credit, and disaster response.
State governments in India are already piloting hyper‑local AI applications, sometimes operating outside the central DPI framework.
Trust and literacy: concrete, demonstrable benefits (e.g., early weather alerts) are prerequisite to building user confidence before large‑scale AI literacy programmes.
Data sovereignty remains contested; distinguishing “government‑critical” data from everyday user data is essential for drafting localisation policies.
AI‑driven credit scoring for informal workers is feasible via analysis of UPI transaction patterns and subsidy histories, but regulatory frameworks are still evolving.
South‑South cooperation (BRICS, AU, Brazil‑India partnership) is viewed as the primary engine for inclusive AI development, complementing multilateral UN initiatives.
Capacity‑building and awareness are the UN’s principal contributions; ultimate governance decisions lie with sovereign member states.

These points encapsulate the panel’s consensus that a development‑centric, inclusive, and collaborative AI agenda is both necessary and achievable for the Global South.

See Also:

India AI Impact Summit 2026

Explorer

ai-for-the-global-south-from-governance-to-inclusion