How Open Networks Are Transforming the Global South

Key Topics Covered

• Open Networks as Economic Infrastructure: Definition, design principles, and distinction from centralized platforms
• Digital Public Infrastructure (DPI): India's UPI, Aadhaar, and ONDC as models for interoperability and inclusion
• AI as a Diffusion Tool: How AI enables complexity-hiding and access at the edges rather than concentrating intelligence at the center
• Trust and Verifiable Credentials: Using data provenance and digital identity (energy credentials, farmer credentials) to enable AI-driven insights
• Global South Implementation: Lessons from India adapted to Indonesia and broader ASEAN adoption
• Sustainable Monetization: Balancing openness, trust, and financial viability without relying solely on philanthropy
• Policy and Governance: Role of government in enabling open network adoption while protecting citizens
• Comparative Economics: Why open networks offer lower transaction costs and different value flows than centralized e-commerce platforms
• Human-Centered AI Design: Ensuring AI reduces cognitive load for end-users rather than gatekeeping or concentrating power
• Practical Use Cases: Energy trading, agriculture supply chains, rural commerce, microfinance, and livelihood creation

---

Key Points & Insights

1. Open Networks Rearchitect Markets, Not Replace Them: ONDC and similar initiatives don't eliminate existing platforms; they create interoperable infrastructure allowing multiple actors (governments, startups, SMEs, civil society) to participate on equal terms. Value flows peer-to-peer, not through a central intermediary.

2. Trust is Built Through Verifiable, Tamper-Proof Data: AI's value in open networks depends on unlocking citizen data and returning it as verifiable credentials (e.g., energy consumption data, farmer identity) that users control and can share with third parties for credit, trading, or services. This shifts from "platform owns data" to "citizen owns and benefits from data."

3. AI's Role is Diffusion, Not Gatekeeping: Successful implementations use AI to hide complexity (three WhatsApp voice notes enable a farmer to trade energy and sell crops across states) rather than concentrate intelligence at a central point. AI should work at the edges where data originates, not only at the core.

4. Decentralized Architecture Prevents Value Capture: Because open networks have no central database and no single platform, they structurally resist the tendency of AI to concentrate power and data. Governance happens at the flow level, not the platform level.

5. Sustainability is Commercially Viable: Open networks can be financially sustainable through real transactions, credit extension, and capital market access (e.g., $1 mutual funds for auto drivers) rather than relying purely on grants. "Commercially sustainable social impact" means profit and inclusion are compatible.

6. Indonesia Model Shows Adaptation from Pioneering to Accelerating: Rather than copy-paste ONDC, Indonesia identified local friction points (17,500 traditional markets needing digitalization, 80,000 cooperatives, 65 million MSMEs facing Chinese competition) and designed open commerce infrastructure (Beckon 2.0) tailored to local needs. This avoids a "social crusader" approach and instead solves real market problems.

7. Scale Without Centralization is Possible at Volume: ONDC processes 7 orders per second with transparent cost of trust that has dramatically declined. The cost of a business delivering a ride on open networks (Namayatri) is ~4 rupees with no commission extraction — 10x cheaper than centralized ride-hailing platforms.

8. Private Sector Leadership is Underutilized: Mastercard's experience bringing a billion people into the formal economy over 11 years demonstrates that corporations can drive inclusion at scale, but this requires shifting from "how" (technology) to "who gets access" (equity). Extending networks to the edges is more important than platform sophistication.

9. Policy Should Enable, Not Prescribe: Government policy works best when it recognizes the opportunity (empowering 70% of the population at the bottom of the pyramid) and creates rules enabling visible livelihood benefits and job creation, rather than protecting incumbent platforms or over-regulating before understanding use cases.

10. AI Models Must Be Built Locally and Contextually: Language learning models (Garia example: 130,000 Indian workers training models in local languages, earning dignified wages) show that rather than exporting Western AI, the most impactful approach is building models "by Indians for Indians" with local workers paid fairly and treated with dignity.

---

Notable Quotes or Statements

• T.T. Koshi (ODC Founder & CEO): "Open networks emerge precisely to address this tension. This is not a technical option but it's a governance choice — a choice which can separate infrastructure from innovation, protocols from platforms and power from participation."

• T.T. Koshi: "Networks do not replace markets, it rearchitect markets. By creating common interoperable rates, they reduce entry barriers specifically for the small business. They enable competition without fragmentation."

• T.T. Koshi: "The digital continent that we dream is not a metaphor. It's a blueprint. It is a call to reimagine digital ecosystem as open, federated and inclusive."

• Sujit Nair (Beckon Labs): "Once you unleash this data which is what API DPI does, it emits the data, gives it back to the user, makes it verifiable and trustable and that enhances the value of AI and therefore AI makes it easier for AI to offer value back to the end user."

• Sujit Nair: "AI's biggest value in open networks has been to reduce the complexity of a shopkeeper... He doesn't know the difference between kilowatt and kilowatt hour. Without that complexity, he said, 'I can earn more tomorrow.'"

• Shamina Singh (Mastercard Center for Inclusive Growth): "The opportunity with the Indonesia model, with the India model, with the open markets model... is to build from the inside out with inclusion at the core."

• Shamina Singh: "AI has to be about benefiting humanity and treating people with dignity. And that's what's really exciting to me about what I'm seeing here."

• Sachin Gopalan (Indonesia Economic Forum): "Open networks give governments a tool to show visible benefits for the people... This is it. Open networks give you that opportunity."

• Sujit Nair (on Namayatri ride-hailing): "The cost of a business delivering a ride... is just four rupees per ride... [and] $250 million of money has gone to the drivers with no commission taken."

---

Speakers & Organizations Mentioned

Speakers:

• Anik Diger — Founder & CEO of Hagdarak
• T.T. Koshi — Founding MD & CEO of ODC (Open Network for Digital Commerce)
• Sujit Nair — CEO of Beckon Labs; Co-founder of Networks for Humanity
• Shamina Singh — Founder & President of Mastercard Center for Inclusive Growth; EVP Sustainability at Mastercard
• Sachin Gopalan — Founder & CEO of Indonesia Economic Forum
• Savvita Mule — CTO at Hagdarak (Moderator)

Organizations & Initiatives:

• Hagdarak — Works at the last mile of the digital economy; network of women agents
• ODC (Open Network for Digital Commerce) — India's open commerce network
• Beckon Labs — Develops open network protocols (Beckon 1.0, 2.0)
• Networks for Humanity — Co-founded by Sujit Nair
• Mastercard (and Mastercard Foundation)
  • Center for Inclusive Growth
  • Global microfinance platform investment ($15M)
• Indonesia Economic Forum — Leading open network adoption in Indonesia
• Garia — Builds language learning models using local workers (130,000+ across India)
• Mandeshi Bank / Rural Women Chamber of Commerce — Piloted on open networks
• Namayatri — Open network ride-hailing platform
• ONDC (Open Network for Digital Commerce, India) — Digital public infrastructure
• UPI (Unified Payments Interface) — Digital public infrastructure example
• Aadhaar — Digital identity infrastructure
• ASEAN — Regional cooperation framework
• Government of India — Ministry of Railways, Digital Economy
• Government of Indonesia — President Prabowo; partnership on digital cooperation (MOU signed Jan 2025)
• Google — Mentioned as booth sponsor/participant
• Uber — Officially joined ONDC ecosystem

---

Technical Concepts & Resources

Architecture & Design:

• Decentralized Architecture: No central database; value flows peer-to-peer
• Interoperability: Ability of multiple platforms/actors to exchange data and transact seamlessly
• Protocol-Based Design: Governance at the flow level, not platform level (Internet banking analogy)
• Unbundling & Interoperability: Key DPI principles driving innovation and competition

Data & Trust:

• Verifiable Credentials: Tamper-proof, user-controlled digital credentials (energy data, farmer identity)
• Data Provenance: Tracking data origin and ownership
• Tokenization: Making credentials portable and tradeable (farmer credentials, utility data)
• Data Ownership Model: Consumer owns data; organization protects it; consumer benefits from it

AI Approaches:

• Agentic AI / Edge AI: AI running at the edges (where data originates) rather than centrally
• Language Learning Models: Local, contextual models trained by and for specific communities (e.g., Garia)
• Complexity-Hiding AI: AI that abstracts technical details so non-technical users (farmers, shopkeepers) can use services intuitively
• Diffusion vs. Supply-Side AI: Focus on distributing AI capability widely rather than building larger central models

Use Cases & Implementations:

• Energy Stack (India): Peer-to-peer energy trading (rooftop solar sales)
• Agriculture Use Cases: Farmer credentials, crop trading across regions (e.g., 300 quintals of onions from UP to West Bengal)
• Open Commerce: Interoperable e-commerce reducing barriers for 75M MSMEs
• Ride-Hailing (Namayatri): 4 rupee cost per ride; $250M to drivers, no commission
• Microfinance Integration: Real-time lending based on visible contracts/transactions
• Traditional Market Digitalization (Indonesia): 17,500 markets × 500–600 sellers each
• Cooperative Digitalization (Indonesia): 80,000 village cooperatives
• Mutual Fund Access: $1 mutual funds for auto drivers and marginalized populations
• Capital Markets Participation: Allowing bottom-of-pyramid populations to invest in upside

Standards & Protocols:

• Beckon — Open commerce protocol (1.0 in India; 2.0 launching in Indonesia)
• MANOV (Accessibility in Indian networks)
• ONDC Protocol — Specifications for open digital commerce
• Curis — Indonesia's QR code payments system (equivalent to UPI)

Metrics & Economics:

• Transaction Volume: ONDC: 7 orders/second; target 50 orders/second
• E-commerce Penetration: India ~8% (B2C); <1% (B2B) despite large platform presence
• Cost Reduction: Namayatri 10x cheaper than centralized ride-hailing
• Financial Inclusion Target: Mastercard: bring 1B into formal economy (achieved 2025); new target 500M people & small businesses by 2030
• Middle Class Growth (ASEAN): 1B joined over last 10 years; 1B more joining over next 10 years; income trajectory $4–5/day → $14–15/day

Policy & Governance:

• Commercially Sustainable Social Impact: Profit + inclusion compatibility
• Proportional Governance: Right-sizing rules to context (not one-size-fits-all regulation)
• Contextual Adaptation: Tailoring models to local problems, not exporting models unchanged
• MOU India-Indonesia (Jan 2025): Digital cooperation framework

---

Document Notes

• Event: India Impact Summit — "Open Networks in the Global South with AI as an Enabler"
• Recording Date: Not explicitly stated; references are to events described as recent (Indonesia open network "recently being launched"; Bharat Vistar/Mahavistar agri networks mentioned as February launches)
• Format: Panel discussion with opening keynote + Q&A
• Audience: Government officials, entrepreneurs, civil society, development professionals