India spends about $150 billion a year on social protection — yet most benefits never reach the people they are designed to support. Tarun Cherukuri founded Indus Action in 2013 to fix that, now partnering with 20 state governments and connecting 3.2 million people to benefits they are legally entitled to but can’t easily reach. His next target: a single-touch delivery system for 800 million people by 2030. Ashoka's Maria Zapata caught up with him to learn more.

Maria Zapata: India guarantees its citizens some of the most progressive social protections in the world — so why aren't people getting them?

Tarun Cherukuri: Think of it as an administrative burden. We all face it — applying for a visa, filing taxes. There’s a discovery problem: people don’t know what they’re eligible for or where to look. Then there’s a documentation problem: proving yourself to the agency trying to give you a benefit. If you’re a pregnant mother, for example, you might fill out a 23-page form and physically visit ten different offices to prove something the health system already knows. And finally, even after you've filed, you're still chasing whether you actually receive the benefit on time and in full.

We’ve documented that accessing an entitlement takes an average of ten burdensome steps — ten! – and the more disadvantaged you are, the more administrative burden you face. The cost of this is not just financial. There’s a deep psychological cost to this experience. And critically, the last-mile government caseworker is equally frustrated. They want to deliver the benefit but can’t. Right now, the whole interaction is designed for friction — sometimes adversarial, always exhausting. What we’re trying to build is something humane and reciprocal, where both people walk away satisfied, and government works for people.

Zapata: India has about 5,000 social benefit programs. How do you choose where to focus?

Cherukuri: The first program we focused on was informed by my own life. I benefited from public scholarships, and I taught in a public school — so I understood education scholarships deeply and started there. But as we followed children, we realized that scholarships only work if maternity entitlements are in place during the first thousand days of a child’s life. And unless parents have access to a secure livelihood — COVID really taught us this — children lose out anyway. So we started mapping the five to ten programs, carefully chosen around life moments that matter most: those with the highest value but also the highest burden and the most difficult to access. If you can ease the process for those five, that becomes the template for every other service as well.

Zapata: A lot of people believe working with government is a dead end. What do you tell them?

Cherukuri: My professor at the Kennedy School memorably called India a "flailing state" — a very strong head, but a weak body. Our plans and policies are robust and rights-based, the product of real social movements. But state capacity lagged behind. That was a helpful frame: it's possible to build state capacity, and that is the order of the day.

What I've learned is that there are no villains here. Everybody's trying to do a good job. The challenge is historical legacy — policies designed for reasons that no longer reflect current realities. What you have to do is find champions within the system at different levels, present evidence, and work closely with them. We've worked across hilly terrains, tribal states, high-capacity and low-capacity states — and universally, it comes down to finding the right champions and walking in with the belief that change is possible.

Zapata: Technology is central to your model, but you insist on keeping community workers in the loop. Why can't tech just do it all?

Cherukuri: Technology can solve information asymmetry problems brilliantly. Google solved discovery. LLMs are solving almost any knowledge question. But entitlement delivery isn't just an information problem — it's a power problem. The interaction between a citizen and a state caseworker is inherently asymmetrical. No matter what the constitution says about equality, the lived experience of that interaction is one of power imbalance. So we always need a human in the loop — to validate the transaction, yes, but also to express care on behalf of the state.

Technology will make that human far more effective. AI can analyze registries and produce a match recommendation to the caseworker: based on what we know about this family, they appear eligible for these five programs. Like a Netflix algorithm, but for social protection. That reduces the discovery burden on the citizen and the caseload burden on the worker. But the human still validates. They're still the face of the state, showing up with care.

Zapata: Tell us about the Universal Entitlements Interface you're building toward 2030.

Cherukuri: India has spent two decades building world-class digital public infrastructure. The identity layer — Aadhaar — gives every Indian a unique digital ID. The payments layer — the Unified Payments Interface — moves money at zero transaction cost, billions of transactions daily. What's missing is a third layer: a validation layer. A construction worker. A pregnant mother. A child in school. A person with a disability. These specific personas need to be validated before the right social protection programs automatically activate.

What we're building stacks all three layers together. Identity, financial, and validation — combined, they enable delivery of entitlements with a single touch. Everything that can be automated, will be. The human caseworker still validates that mother, that child, that worker — but the maze of ten burdensome steps collapses. Once we pressure-test this across a significant number of states, across geographies with very different levels of digital access and literacy, we want to open-source those protocols. If it works here, it can work anywhere.

Zapata: Tarun, how do you see the moment we are in, especially with AI — and what’s ahead?

Cherukuri: We're in a revolutionary moment. With every revolution, there are risks, and these AI risks are real. But there's also a huge opportunity to channel AI for public good. I take confidence from the long arc of history — it bends toward justice. Coming from India, we're products of a freedom movement that taught us truth and nonviolence are, in the long run, the most powerful instruments. I try to apply that daily. The latest technological breakthrough is just the newest instrument — and I'm optimistic we can point it toward the people who need it most.

Tarun Cherukuri is the founder and CEO of Indus Action and an Ashoka Fellow . In April 2026, he accepted the Skoll Award — one of social entrepreneurship’s most prestigious honors — in Oxford.