The futures of India Stack
Translated from the French, which remains the reference text. This translation is under review.
Buoyed by its success, India Stack has become an important instrument of Indian Soft Power, and the potential bedrock of a technological "third way".
It is developing, on the basis of the open-source, secure and decentralised approach that is its hallmark, in three major directions: towards the Global South, with the benevolent support of the UN and the World Bank, through MOSIP in particular (6.1); towards the private sector, by decentralising e-commerce through the Beckn protocol (6.2); and lastly in a systemic project, carried by Networks for Humanity, which extends the same reasoning to networked finance, intelligence and trust and applies it to critical infrastructures, starting with energy (6.3).
6.1MOSIP, CDPI and the export of India Stack (India's soft power)
Faced with closed global platforms, India is actively exporting its DPI model to developing countries. The aim is to spare those countries from reinventing the wheel or buying costly systems (Vendor lock-in).
"Now, the question is: globally, how can as many countries as possible build and deploy their own DPI building blocks without taking decades and spending billions of dollars? [...] Countries that choose to be fiscally prudent would be wise to leverage shared digital infrastructure and drive reuse, efficiency, and interoperability rather than rebuilding full-stack solutions each time. Full-stack solutions, in most cases, tend to result in the creation of proprietary systems that are not in the interests of countries when it comes to critical digital infrastructure."
— Dr Pramod Varma, Carnegie Endowment article
Through initiatives such as CDPI (Centre for Digital Public Infrastructure) or the open-source identity platform MOSIP, India is already helping dozens of countries (in the Philippines, in Africa) to build their own digital sovereignty on the foundations of India Stack.
The vision
A modular stack, open and sovereign by default
Professor S. Rajagopalan (chair of MOSIP) and Dr Pramod Varma (chief architect) describe the technological captivity (vendor lock-in) that prompted the creation of MOSIP in 2018:
"Many countries initially tried the conventional route: write requirements, issue tenders, pick a vendor and buy a proprietary solution. That often led to vendor lock-in, high costs and limited govt control over citizen data. Most of the nearly 20 countries that attempted this ended in lots of trouble. The vendor would say, 'I'll give you identity, per identity, so many dollars'."
— S. Rajagopalan and Pramod Varma, via The Times of India
MOSIP's guiding principle lies in the "M" of its acronym: Modular. Unlike Western monolithic systems, MOSIP is built like a set of Lego bricks (an API-first micro-services architecture).
A state downloads the software free of charge (via GitHub), hosts it on its own national servers (Sovereign-by-design), and is free to buy its biometric scanners from supplier A and its physical cards from supplier B, thereby avoiding the monopoly.
"By using open-source software and standardized interfaces, MOSIP enables governments to select and change technology providers without being constrained by proprietary systems or biometric templates. [...] Data creates power. Countries must retain control over the data generated through digital systems. This is why sovereignty considerations are central to MOSIP's approach."
— Architectural doctrine, MOSIP technical documentation
The global academic community quickly nicknamed MOSIP "Aadhaar in a box".
"To facilitate Aadhaar's replication in other countries, its proponents have supported the creation of an open-source platform known as MOSIP (Modular Open Source Identity Platform)—colloquially referred to as 'Aadhaar in a box'—which any country can implement license-free. MOSIP aims to help implementing countries ensure uniqueness in their identification systems, achieve interoperability, pursue privacy-by-design while reaching scale, avoid vendor lock-in and maintain affordability."
— Academic journal ResearchGate, 2020
The strategy: digital non-alignment and neutrality
After Aadhaar's phenomenal success in India (more than a billion people enrolled), many developing countries, trapped by Western companies, approached the Indian government and Nandan Nilekani to replicate that "miracle", particularly useful for targeting the distribution of social benefits.
Nandan Nilekani and the team reached a clear conclusion, however: Aadhaar could not be exported as it stood. The code had been designed specifically for Indian infrastructure and, above all, it belonged to the state (UIDAI).
Aadhaar's concepts had to be rebuilt from scratch, in the form of a free and open-source Digital Public Good.
"Aadhaar itself could not be shared, but its core ideas could be rebuilt as an open digital public good for the world. [...] Over the years, it became clear that many of India's learnings were globally applicable. Around the world people want to use digital transformation to change their societies. And you need to have a way of doing it which is low cost, high performance, and manage even small transactions."
— Nandan Nilekani, The Times of India
Digital non alignment
Through MOSIP, India positions itself as the benevolent technological leader of the Global South, offering a "Third Way" as an alternative to the monopolies of the American GAFAM and to the surveillance infrastructure sold by China.
Academic and diplomatic circles have coined a term for this strategy: "Digital Non-Alignment", a historical nod to Nehru's diplomatic doctrine during the Cold War.
"India has carved out its own path through its ‘Digital Non-Alignment’ strategy. By adopting the spirit of the original Non-Aligned Movement, this strategy allows developing nations to harness digital technology without losing sovereignty over their data and governance structures. Instead of exporting proprietary solutions, India has come up with open-source models that can be adapted based on the social and legal needs of individual countries. [...] Politically, India projects itself as a South–South partner offering technology over debt, positioning its open-source digital stack as a trusted alternative to China’s Belt and Road model."
— Grimshaw Club think tank, London School of Economics
The combination of the expertise of figures such as Nandan Nilekani and Pramod Varma, of the academic world (IIIT-B) and of international philanthropic funding has allowed Indian civil society to achieve a feat: making its architectural philosophy (India Stack) a de facto quasi-standard for digital public infrastructure across the global South.
Neutrality
Indians have grasped a fundamental geopolitical rule: if the code were the property of New Delhi, foreign states would refuse to adopt it for fear of interference or cyber-espionage. To guarantee absolute neutrality, MOSIP's governance was entirely decoupled from the Indian state.
The academic anchoring
MOSIP is not a government agency. The project is incubated, developed and hosted at IIIT-B (International Institute of Information Technology, Bangalore), an independent research university.
Sharad Sharma, founder of the iSPIRT think tank, reads this strategy of institutional detachment as the decisive key to MOSIP's success:
"There is much to learn from MOSIP in this regard. Today, it is the most successful global Digital Public Good from India. [...] MOSIP is based in IIIT Bangalore, and its institutional structure has been instrumental to its successful globalization. Why does this matter? Take Electronic Voting Machines (EVMs) as an example. Despite their extraordinary success in India, they haven't been adopted by any other country. The main reason is that the creation and governance of public-tech are in the hands of the Govt. of India, and this worries other Governments. MOSIP breaks this jinx."
— Sharad Sharma, official iSPIRT archives
Global philanthropic funding
To establish its international legitimacy, MOSIP is funded by a consortium of major philanthropic foundations rather than by the Indian taxpayer.
"Initial development was supported by philanthropic funding from organizations including the Bill & Melinda Gates Foundation, Omidyar Network [created by eBay's founder] and Tata Trusts. [...] The International Institute of Information Technology Bangalore (IIIT-B) established MOSIP in 2018 to be a core component for Digital Public Infrastructure and hosts the initiative."
— Institutional quotation, Biometric Update / IIIT-B
Adoption to date: the new standard for identity worldwide
MOSIP's expansion now rests on a powerful institutional structure. The World Bank (through its ID4D initiative) has become its principal sponsor, recommending the tool to countries requesting loans for their digitisation.
The deployment figures, updated at the end of 2024, are staggering. MOSIP is establishing itself as the default sovereign infrastructure of developing countries. According to MOSIP's official 2024-2025 progress report, the global footprint counts 29 countries in a deployment, pilot or study phase (19 of them in Africa, but also in Latin America and Asia-Pacific), covering a potential population of 970 million inhabitants; actual adoption, for its part, exceeds 185 million citizens who have already received their digital identity through a MOSIP-based system.
Among the major use cases, according to the same report, the Philippines (PhilSys programme) are the largest operational success in the world, with close to 94 million identities created (98.6 % of the population enrolled). Ethiopia (Fayda ID programme) counts more than 33 million people registered (32 % of the target); notably, it uses the MOSIP architecture in collaboration with the UN refugee agency to register refugees and open access to bank accounts for them. Morocco (National Population Register) is the very first country to have signed (in 2018), with more than 24 million people enrolled (88 % of the population), completely restructuring its social benefits system. Togo, finally, has seen an extremely rapid deployment, with more than 3.5 million citizens registered, used for emergency cash transfers.
6.2With and after DPI: Beckn and e-commerce
"DPI can be used to create open source networks that “unlock” the relationships between buyers and sellers, allowing both Amazon and your local bookseller to exist on the same playing field. [...] Because the same protocol can be used for many purposes, services using open protocols are interoperable, meaning someone could access a mobility solution at the same time that they order food—from two different parties. [...] DPI is a 'digital highway' that provides the means for all sorts of parties—entrepreneurs, consumers, nonprofit organizations—to connect on a level playing field."
— Dr Pramod Varma, talk at the Yale MacMillan Center
The project: the Open Network Digital Commerce (ONDC)
"The ONDC enables e-commerce to shift from a platform-centric model to an open network model, ensuring greater discoverability, transparency and interoperability between players. A product listed on any one platform (e.g. Myntra) would be discoverable and available for purchase from any other platform (e.g. Amazon, Flipkart), improving visibility and market reach for players. [...] Practically speaking, ONDC is taking the four steps of commerce that were traditionally taken up by one entity: discovery, assortment, fulfillment, and governance and unbundling it so that multiple players can execute each of these steps independently."
— Sattva Consulting report / ONDC strategy document
"Three big revolutions will be—democratizing credit through AA [Account Aggregator Framework], democratizing commerce through ONDC; and essentially making the delivery of products much simpler and cheaper with logistics' transformation. These three steps will lay the foundation for equitable commercial activity for both goods and services. India will move from a pre-paid cash informal economy to a post-paid formal cashless productive economy."
— Nandan Nilekani, Livemint interview
ONDC is the Indian state's answer to the duopoly of Amazon and Flipkart (owned by the American company Walmart), which controls the vast majority of Indian e-commerce. The starting observation is alarming: although India has some 60 million micro-enterprises (MSMEs) and small neighbourhood shops (Kirana stores), barely a fraction of them manage to sell online, stifled by commissions ranging from 18 to 40 % and by the opacity of the algorithms.
The vision: breaking the silos and democratising the market
ONDC's founding vision is not to create a "state Amazon" but to invert the economic model of the internet. ONDC seeks to unbundle the value chain (discovery, ordering, payment, logistics) and turn it into an open network.
Thampy Koshy, the chief executive of ONDC, explains the absurdity of the current captive model and the restorative vision of the network:
"Today, only a few platforms have held the market share for e-com globally and for each of them, if you are a seller on Platform A, you can only sell to those buyers who are registered on that platform. [...] Because buyers and sellers can only communicate if they are on the same platform, therefore, today nobody can set up a 'buyer only' or 'seller only' platform, constricting e-commerce. [...] With ONDC, a big change is coming - anyone who has digital consumers can integrate itself with ONDC as a buyer platform and will be exposed to a bigger number of sellers."
— T. Koshy, CEO of ONDC
ONDC aims to replicate the "UPI miracle" (which unified payments) in the physical world. Nandan Nilekani sums up this approach from first principles:
"India is on the cusp of a digital transformation. [...] And yet digital commerce in India is just 7 percent of the overall market, constrained by various barriers. [...] It is founded on the core principles that digital services could be 'presence-less', 'paperless', 'cashless', and 'consent-based'. All the components of India Stack were built using a first-principles approach... to solving for economic primitives."
— Nandan Nilekani, Democratising Digital Commerce in India
The technical architecture: the Beckn protocol
"Beckn protocol creates a common language for interoperability among different digital platforms for facilitating economic transactions, similar to HTTP for the internet or SMTP for emails. Beckn is technology and sector-agnostic, enabling all aspects of digital commerce transactions, from discovery to post-fulfillment activities. [...] By unbundling the services and allowing each stakeholder to specialise, the network ultimately reduces dependencies on any particular platform and fosters collaboration, innovation, and inclusivity."
— Technical explanation, official Beckn report / Publicis Groupe
In brief
To understand how this commercial unbundling works, the technical layer (the protocol) must be distinguished from its implementation at the scale of India (the institutionalised marketplace). The protocol, Beckn, is a common computing language, open-source and agnostic: just as the HTTP protocol allows any browser to display any web page, the Beckn protocol allows an application A to talk to an application B (for discovery, ordering and fulfilment) without the two having signed a bilateral contract. The network, ONDC, is the practical national initiative (backed by the Ministry of Commerce) that uses the Beckn protocol to build the Indian e-commerce network in concrete terms, in order to disrupt the Amazon and Flipkart duopoly.
On ONDC, the monopoly is broken end to end: a customer can search for vegetables from a client application (Buyer App), find the produce of a small grocery using third-party software (Seller App), and be delivered by an independent logistics company (Logistics App).
Governance: a hybrid form (Section 8 Company)
ONDC was launched at the end of 2021 under the aegis of DPIIT (Department for Promotion of Industry and Internal Trade, Ministry of Commerce). To guarantee its agility and its neutrality, however, ONDC was structured as a Section 8 Company (a private not-for-profit company).
The seed capital was raised from a consortium of banks (State Bank of India, HDFC, Kotak Mahindra) and institutions such as BSE and NABARD. The Advisory Council brings together Nandan Nilekani, Dilip Asbe (chief executive of NPCI) and Adil Zainulbhai (Quality Council of India). The state gives the impetus, the private sector executes, and the absence of a profit motive guarantees that the infrastructure will not tax sellers unduly.
Adoption, industrialisation and spread: the test of reality
While ONDC's architecture and the Beckn protocol may at first have looked like brilliant theoretical experiments, the adoption data for 2023-2024 show that unbundling has passed the industrialisation barrier (Scale). ONDC demonstrates that an open protocol can generate massive network effects without depending on a proprietary monopoly.
Domestic hyper-growth: the surge in ONDC volumes
Adoption of the network has accelerated exponentially, confirming the relevance of public infrastructure in a fiercely competitive market. While the network was managing a few tens of thousands of transactions in early 2023, it passed the mark of 12 million transactions a month by mid-2024, carried jointly by retail and by mobility. By the end of 2024 the network had passed 600,000 connected sellers, spread across more than 1,200 towns and cities. The wager on economic inclusion holds: more than 80 % of these merchants are small or micro-enterprises (Kirana stores and SMEs) that had often never sold online because of the stifling commissions (ranging from 20 to 30 %) imposed by the traditional e-commerce giants.
The breakthrough use case ("Killer App"): mobility with Namma Yatri
This is the Beckn protocol's most striking and most political success. Tired of the opaque algorithms and the commissions of Uber and its Indian competitor Ola, a union of tuk-tuk drivers in Bangalore created Namma Yatri, a decentralised ride-hailing application built on ONDC's rails. The application works on a "Zero Commission" model: the customer pays the driver directly (in particular through UPI). The outcome is a textbook case of platforms being disrupted by the digital commons. By the end of 2024 Namma Yatri had passed the mark of 100 million rides, generating tens of millions of dollars of direct, untaxed income for hundreds of thousands of drivers. The model is now spreading across several Indian states, proudly nicknamed the "Linux of mobility".
International spread: the Beckn protocol as diplomatic "Soft power"
Aware that its architecture acts as an agnostic language (a lingua franca of commerce), the Indian ecosystem no longer confines this model to its own borders. Beckn, maintained by the independent FIDE foundation (chaired by Nandan Nilekani), is becoming an instrument of geopolitical export.
Cross-border B2B commerce, first: ONDC has officially opened its network internationally. The Singaporean platform Proxtera (supported by the Monetary Authority of Singapore) has become the network's first international "Buyer Application". Through this interoperability, a buyer in South-East Asia can browse the catalogue of an Indian factory, negotiate and pay in real time, with the protocol standardising the whole cross-border commercial chain.
Digital Non-Alignment towards the Global South, next: in 2024 Indonesia officially launched the Indonesia Open Network, a digital commerce network built entirely on ONDC's "blueprint". Similar projects are emerging in Latin America (the Belém Aberta network in Brazil) or in Africa (Open Gambia Network), giving those countries immediate software sovereignty with which to digitise their economies, without falling into the "vendor lock-in" of Western platforms.
Europe's relative interest, and the bridge to the data spaces: the Beckn protocol is beginning to attract interest in Europe. Through the Open Commerce Bridge (OCB) project, funded by the European Commission's NGI Sargasso initiative, the Startin'Blox cooperative has worked to merge the Beckn protocol with the standards of the semantic web and those of the European data spaces. This specification, named Beckn-LD, has the explicit aim of making Indian decentralised transactions natively compatible with the ecosystem of the future data spaces, with successful validations in short-supply-chain local agriculture and in media exchange.
Multi-sector expansion, finally, towards the internet of energy: as evidence of the protocol's agnosticism, Beckn is moving beyond mobility and commerce. Recent initiatives are taking it up to build the Unified Energy Interface (UEI). The aim is to use this language of interoperability to unify the fragmented energy ecosystem (electric-vehicle charging, for instance), allowing any application to discover and pay for any charging point in the country, and so tracing a path towards a global Digital Energy Grid.
The blind spots of unbundling: the warning from the Internet Freedom Foundation (IFF)
The revolutionary architecture of Beckn and of ONDC nonetheless creates new structural blind spots, brought to light by the NGO Internet Freedom Foundation (ONDC: An Explainer). By unbundling services, the network mechanically dilutes the chain of responsibility.
The first blind spot is the gap in legal liability (grievance redressal). On an integrated platform (such as Amazon), the consumer knows whom to contact for a refund if a product is faulty. On the ONDC network, if a customer orders through App A a spoiled meal prepared by Seller B and delivered late by Logistics provider C... who is legally liable? With no central platform, the consumer risks meeting a chain of deflected responsibility between the connected applications.
The second is the proliferation of private data. To execute an unbundled transaction, the buyer's sensitive information (name, address, basket) has to be passed by the Buyer App to the Seller App and to the Logistics provider. The IFF is concerned about the consequences of this dissemination of data across a multitude of third-party servers, which sharply increases the attack surface for non-consented profiling, particularly in a country where the legal framework for data protection (DPDP Act) is still recent.
6.3The NFH and the extension of the paradigm at the heart of Beckn and of India Stack
"Together, Finternet and Beckn, powered by an underlying credential and trust infrastructure, form a common ground [...], where anything of value can be represented, discovered and exchanged [...]."
— Networks for Humanity, nfh.global, "About us"
In brief: what is the NFH?
For some years now, the architects of India Stack have no longer worked for the Indian State. Pramod Varma said so at the outset in Paris, on 16 September 2026, at Inria: "I have no connection with the government of India"; what the team learnt by building systems at population scale, it now carries in a structure named Networks for Humanity (NFH), created in early 2025, which defines itself as "an international network of frontier labs building open and interoperable systems for the agentic economy".
The NFH brings together four entities: FIDE, formerly the Beckn Foundation, is its Indian entity and signs its publications; three entities created in Singapore, the United States and Switzerland extend it. The labs are in Bangalore, Singapore, Zug (Switzerland) and San Francisco; the funds come from Nandan Nilekani's philanthropy, the Gates Foundation, Google.org, the Asian Development Bank and Spectrum Impact; Nandan Nilekani is its co-founder and patron. The intellectual property is placed, according to Pramod Varma, under the Swiss entity, for the same reason that took MOSIP out of the State: not to be geopoliticised. The governance of the whole is not public for the time being. From Aadhaar to the NFH, the reasoning is the same: design minimal protocols, entrust their governance to a guardian who does not own them, and let a multitude of actors build and operate as a network (schema 6.1).
Schema 6.1 · From India Stack to Networks for Humanity : a thin protocol, a multitude of operators
The NFH's project
Its purpose holds in three words: Networks for Humanity.
It can be summed up in one sentence: a systemic project articulating networked exchange (Beckn), networked finance (Finternet) and networked AI (PAC), through an engineering of trust in a distributed setting, in order to transform the critical systems of our societies, starting with energy infrastructures.
We set out below the main components of this project, except Beckn, already presented in the previous chapter.
Finternet: networked finance
The observation. Today's finance rests on a chain of intermediaries bound by bilateral contracts, in paper form and imposing checks repeated at every step. Hence high costs, fraud and slow services. A large share of people and small businesses remain excluded. This model prevents the network effects that made the strength of the Internet.
The proposal. To do for finance what India did for identity and payments (Aadhaar, UPI), and for the exchange of goods and services (Beckn): common "rails", open to all, on which actors compete. The core of the project, presented to the Bank for International Settlements in April 2024 and then detailed in the Finternet Labs white paper, is tokenisation: every asset, and every right attached to it, becomes a programmable digital token, exchangeable directly between two parties.
Six innovations. Every asset becomes mobilisable, whether created by the user (credentials, personal goods), attested by a third party (a painting authenticated by a gallery), registered (land, vehicles) or regulated (money, securities), and combinations become possible: the authenticated painting secures a loan. Ownership is unbundled, with use, rental, building rights or pledge circulating separately, and an expensive asset is fractionalised. The user takes back control: they choose their asset manager, gather their holdings in a single account, set their own rules and decide which data they disclose. Compliance follows the operations rather than barring the entrance: authorities set their rules per flow, who may take part, for which asset, up to what amount, and the system blocks in advance whatever is not compliant. The network replaces the bilateral contract: everyone deals with everyone within a common framework, with no platform centralising. Supervision, lastly, moves from the sample to the proof: every operation leaves a verifiable trace, the fight against fraud becomes continuous, and disputes would go to online arbitration services.
For whom, and in what spirit. The examples target those whom finance leaves aside: the self-employed, artists, small businesses, agricultural markets, targeted aid (health or education vouchers), and climate finance (green bonds, carbon credits, subsidies). The Finternet takes from cryptocurrencies the token and programmability; but it does without mining and without a currency of its own, and it brings in regulated assets and regulators. It presents itself as an evolution of the existing system, not an alternative (schema 6.2).
Schema 6.2 · The purpose leaves the rail for the object
PAC: networked AI
Project PAC, presented in Paris in September 2026, extends to computing and intelligence the approach at the heart of DPI, of Beckn and of the Finternet. Its aim is to let actors draw on the intelligence of the network. The acronym reads two ways: p for "personal", because each node works for an individual or a small business who keeps control of it; P for "planetary", because the system is designed to work at very large scale.
PAC is presented as "a planetary scale networked autonomous computing system of agents". Each node of the network acts on behalf of a person or an organisation. Data stay with their holder, protected in place. The agents exchange and conclude transactions through techniques of joint computing between agents: several parties obtain a result together without showing each other their data. The possibilities put forward are of this order: a patient's assistant that would submit their records to every specialised medical agent in the world to obtain the best diagnosis, without any agent seeing the raw data; several companies that would build an optimal logistics plan together without disclosing their strategic data; several institutions that would detect fraud together at network level, without any system centralising all the data; people who would connect without entrusting their data to a central platform, enough to rethink social networks. These are possibilities, not achievements.
The project rests on a wager: AI models, the capacity to execute and creation will become abundant, personalisation will happen at the edge, on the user's side, value chains will be unbundled, agents will become the new customers of businesses, cryptography will found new forms of guarantees, and only openness will allow true diffusion: "Network will be the real AGI".
The trust framework
The open networks the NFH promotes give up the central platform which, in the dominant model, vouches for the participants. This is the case of Beckn for the exchange of goods and services, of the Finternet for the circulation of assets, and of PAC for computing shared between agents. Trust must therefore come from elsewhere: from proofs that travel with the exchanges and that anyone can verify without asking a third party. The NFH entrusts this function to two complementary building blocks. OpenCred produces and verifies the proofs. DeDi publishes the registries that tell whom to trust.
OpenCred: the proof. OpenCred is free software (MIT licence), published by NFH Trust Labs, for issuing and verifying verifiable credentials to the W3C standard (Verifiable Credentials). It runs on the issuer's computer or on their own server, and the signing keys never leave the issuer. The holder keeps the credential and chooses to whom to present it; the verifier checks its proof without contacting the issuer; if OpenCred disappeared, everything issued with it would remain verifiable. It is aimed first at small issuers, the school that delivers transcripts, the SME that draws up employment certificates, and almost any statement can become a credential: diploma, prescription, professional licence, identity document, credit certificate, origin of a product, ticket, supplier approval, down to the receipt.
DeDi: the directory. DeDi (Decentralized Directory) answers the questions the credential alone does not settle. Is this issuer entitled? Is this key really theirs? Has this document been revoked? DeDi is a protocol for publishing public registries, organised in namespaces, registries and records: the list of banks licensed by a central bank, an issuer's public keys, the list of revoked credentials. Each publisher signs and publishes its own files; DeDi servers index them without standing in for them, they are not authorities. The protocol comes under LF Decentralized Trust, within the Linux Foundation, and the NFH runs a public and free instance of it, dedi.global.
Two building blocks, four questions. OpenCred checks that the data are intact, which key signed them, whether that key can be trusted and whether the credential is still valid. The credential answers the first two questions; the last two refer to DeDi's registries, the issuer's trust chain and the revocation status. Trust is graded there: an issuer attached to a national authority by a signer certificate, on the model of electronic passports, an issuer in the course of attachment, an issuer who publishes its own keys. This framework therefore creates no authority; it makes existing authorities machine-verifiable and lets those who come under none issue.
What the framework brings to each floor, in the study's reading. In Beckn, the function of admitting and controlling sellers, which the platform assumed, disappears with it; the framework can rebuild it in a distributed way: approvals presented as credentials that the buyer's application verifies before ordering, the list of participants published in DeDi, the receipt or proof of delivery becoming itself a credential reusable for a loan or a dispute. In the Finternet, each token carries a "chain of proofs", credentials and history travel with it, and the next actor does not have to re-verify everything; the discovery and participant-registry functions, which the architecture provides for without naming them, are the place DeDi can take. In PAC, OpenCred proves what one chooses to disclose while PAC computes jointly without disclosing anything; the articulation is not yet written, but the examples call for it, a registry of entitled medical agents, the proof that an institution is a licensed bank, the mandate of an agent acting for a person. Credentials would thus supply the inputs and the entitlements of joint computing.
Digital Energy Grid (DEG): the application to energy infrastructures
In February 2025, FIDE published, with two chapters by the International Energy Agency (IEA), a white paper, Digital Energy Grid — A vision for a unified energy infrastructure, with forewords by Nandan Nilekani and by Fatih Birol, the IEA's Executive Director; the Agency contributes to it without making it its own view. Energy becomes the first field where the NFH's building blocks are assembled.
The observation. The electricity system is changing under three pressures: decentralisation, digitalisation and rising demand. Rooftop solar could reach nearly 3,700 GW by 2035, and electricity demand could nearly double by 2050. Yet existing solutions are specialised, siloed and often still on paper: an equipment certification cannot be verified digitally, an energy transaction leaves no usable trace. The paper draws from this three coordination challenges, bearing on the flows of information, of transactions and of energy.
The proposal. To double the physical grid with a data network. Like the Finternet, it is a "network of networks": each node (microgrid, aggregator, home) remains autonomous but interoperable. Three building blocks: a universal identity for each asset and each actor, on the model of telephone dialling codes; machine-readable data rather than PDFs; verifiable and portable data, signed, shared with consent. The DEG "is not a product": it is a base layer that connects the sector's existing standards without replacing them.
The layers brought together. The white paper's central scenario brings them all into play. Mihir, a farmer, wants to install subsidised solar panels. Today he files three applications (subsidy, loan, grid connection) and must each time prove his identity, his ownership and his eligibility. With the DEG, a single step suffices: installers are found and engaged on the strength of their verifiable credentials, that is Beckn and the trust framework; the bank verifies eligibility for the subsidy directly; production data are shared with Mihir's consent; the subsidy is paid automatically as soon as the installation is verified, that is the logic of the Finternet. Other scenarios describe demand response settled on signed meter readings, the rental of home batteries, energy exchange between neighbours, a global market of tokenised carbon credits with "no double counting", and the fractional ownership of shared renewable plants, for which the white paper cites the Finternet. It describes, lastly, an "agentic energy ecosystem": a home assistant is asked to cut the bill, and it concludes micro-contracts on its own; business agents "verify credentials, and negotiate contracts". That is the world PAC means to equip.
From vision to implementation. The DEG advances by progressive adoption, relying on the Unified Energy Interface (UEI), a Beckn network for electric-vehicle charging, peer-to-peer exchange and demand response. Several steps have followed: the DEG specification defines energy credentials in the W3C format (green energy, ownership, eligibility for aid, enrolment in an exchange programme), whose revocation goes through DeDi; India's Ministry of Power is preparing an India Energy Stack, whose version 0.3 appeared in February 2026 for a completion targeted for July 2026; at the end of September 2026, Schneider Electric and the NFH signed a memorandum of understanding to explore, through pilots, applications of the DEG in India: industrial and commercial demand response, home energy management, fast charging, carbon accounting. Energy is thus, in the study's reading, the first sector where the NFH's four layers, exchange, value, trust, intelligence, converge in a single architecture (schema 6.3).
Schema 6.3 · The digital energy grid
Conclusion
What the NFH proposes is to extend and complete the edifice begun with India Stack, continued with DPI and with Beckn. Identity, signature, payment and data sharing made a billion people full participants in the digital economy; Beckn opened the exchange of goods and services outside the platforms; the Finternet means to do the same for assets, PAC for computing and intelligence, and the trust framework supplies the piece that was missing on every floor, the proof verifiable without a third party. The reasoning has not varied since Aadhaar: a minimal protocol held in common, a multitude of actors who build and operate, a guardian who does not own. What has changed is the scale, an edifice that now means to be planetary, and the nature of the guardian: no longer the Indian State, but a network of labs on three continents, funded by philanthropy, whose governance remains to be made public. The wager is that the agentic economy requires this open infrastructure, failing which it will concentrate in a few platforms of the new era.
In many respects, this project meets Europe's. The verifiable credentials are those the European identity wallet adopts; DeDi's signed registries answer the need covered, on our side, by the eIDAS trusted lists and the participant registries of the data spaces; the tokenisation of assets meets the European pilot regime for market infrastructures on distributed ledgers; the DEG speaks of the same objects as the smart grids and the interoperable charging that Europe regulates; and the Beckn-LD bridge towards the data spaces (6.2) already exists. Both sides pursue the same end, open and interoperable networks where no single actor holds the key, by the same method, trust in a distributed setting. They have not sought it by the same paths: India delivered protocols and implementations, then let the market and the law follow; Europe wrote the law and the standards before the implementations, and few of its protocols are in service, as Henri Verdier remarked when hosting Pramod Varma. Nor does it hold the guardian in the same way: the European foundation for digital public infrastructure announced by the open source strategy (3.4) answers the question the NFH leaves open, that of a guardian whose governance is public.
It is this gap, not of purpose but of sequence, of method and of governance, that the third part takes as its object. Having gone through the two corpora each in its own logic, we set them side by side: a crossed and comparative analysis, which looks in the one and in the other for the same regularities, the same tensions to resolve and the same operators for moving from tension to solution. From this reading emerges a pattern language: not a model to copy, but a repertoire of proven forms, each named and embodied on both sides, which European and Indian actors can pass from hand to hand to build, together or separately, open, interoperable and decentralised infrastructures.