How Digital ID Systems Make Refusal Impossible

Key Takeaways

• Digital identity systems often launch as voluntary tools and quietly become de facto mandatory once tied to essential services.
• India's Aadhaar enrolled over a billion people, was repeatedly linked to large-scale data breaches, and became practically impossible to refuse.
• The EU's Digital Identity Wallet is framed as optional today, but history suggests that framing rarely holds once adoption reaches critical mass.
• The people most harmed by digital ID failures are typically those with the least power to challenge or correct them.

Tomorrow, July 11th, is World Population Day, the UN's annual occasion for thinking about what it means to have over eight billion people sharing a planet and the systems built to account for them all. I want to use it as an opportunity to discuss a specific kind of system that is becoming more universal by the year: digital identity. Not because it’s inherently bad – it often starts as a genuine attempt to solve real problems of exclusion and access – but because at the population scale, something fundamental changes about what it means to opt out.

From Civil Records to Digital Infrastructure

Not everyone on this planet has a legal identity. The UN estimates that around one billion people worldwide lack basic identification documents. Across much of sub-Saharan Africa, large portions of the population were never formally registered at birth, which means no birth certificate, no passport pathway, and reduced access to healthcare, banking, voting, and public benefits. The same applies to China and second-born children under the country’s one-child policy. 

UNFPA and national governments have been working together through Africa's Civil Registration and Vital Statistics programs on the premise that legal identity is foundational to inclusion in public life. That is a reasonable premise.

This context matters because it’s part of how digital identity systems get sold: as a solution to exclusion. For people who have no documentation at all, gaining formal recognition can be genuinely transformative. I don't want to dismiss that. The problem is what tends to happen after a system is built to include everyone and then becomes the single gateway through which everyone must pass to access anything.

When “Optional” Loses Its Meaning

India's Aadhaar is the clearest case study of this dynamic at the population scale. Launched in 2009 as a voluntary biometric ID program, it enrolled over a billion people and was gradually linked to bank accounts, mobile phone registrations, tax records, and welfare transfers. 

Courts ruled in 2018 that mandatory linkage to private sector services was unconstitutional, but by then the system had already become embedded deeply enough in public service delivery that refusing it meant opting out of significant parts of civic and economic life. The "voluntary" label was technically accurate and practically meaningless.

The centralization that made Aadhaar powerful also made it dangerous. Freedom House's research documented that in the years since the system's widespread adoption, breaches of third-party databases holding Aadhaar data resulted in leaks of hundreds of millions of records, fueling a black market for fraud and cybercrime. Building a single unified identity infrastructure for a billion people creates a single point of failure at a scale that is genuinely hard to comprehend. When that infrastructure is breached, the consequences aren't a privacy violation for one person. They’re a systemic event.

The European Union is now building its own version. The EU Digital Identity Wallet, being rolled out under the updated eIDAS 2.0 regulation, will allow citizens to store verified credentials on their phones and present them across public and private services throughout the EU. The official framing is that the wallet is voluntary. Member states are required to offer it, but no citizen is required to use it. 

Aadhaar was framed the same way at the start. The moment hospitals, employers, banks, and government agencies adopt the wallet as their preferred method of verification, the practical cost of not having one starts to rise. Voluntary systems tied to access tend to stop being voluntary, not through a single policy change, but through an accumulation of small decisions that seem reasonable in isolation.

Who Controls Access to Identity

The other dimension is who sits at the center of the infrastructure. Privacy International's long-read on identity gatekeepers and the future of digital identity examines how both governments and private companies are positioning themselves as the authorities who verify, grant, and, if necessary, revoke identity in digital spaces. When those gatekeepers decide your credentials are invalid, expired, or disputed, your ability to access services goes with them. 

This isn’t a theoretical risk. People in countries with functioning legal systems have had benefits suspended or access denied because of errors in digital ID databases, cases of mistaken identity, or bureaucratic failures that took months to correct because adequate human review processes were not built alongside the automated ones.

At the population scale, even a low error rate produces an enormous absolute count of people affected. A system with 99.9% accuracy and one billion users still fails a million people. Those failures aren’t distributed randomly across a population.

Population Scale Changes the Stakes

I support making legal identity accessible to people who have been excluded from it. I support building digital tools that reduce friction in interactions with public services. What I can’t support is the idea that rolling out digital identity at a population scale is automatically safe, or that the "optional" framing governments use at launch accurately describes how systems function after mass adoption.

The history of large-scale digital identity programs shows a consistent pattern: voluntary rollout, growing linkage to essential services, rising practical cost of refusal, and a centralized data infrastructure that becomes an increasingly attractive target for both state and criminal actors. 

Privacy and autonomy don't disappear in a single dramatic moment. They erode incrementally as each new service integration makes opting out slightly more costly, until the cost becomes prohibitive for most people and the system is no longer meaningfully optional at all.

Eight billion people is a useful frame for this because it’s the actual scale at which these decisions are being made. Digital identity infrastructure built for a planet of eight billion people and designed primarily for efficiency, inclusion, and convenience, without equivalent investment in the safeguards that protect privacy, correct errors, and prevent coercion, isn’t a neutral tool. It’s a system of control that most people won’t have a chance to refuse. That’s worth naming, and worth getting right before the architecture becomes too entrenched to change.