When Inclusion and Control Look the Same: Biometrics, Power, and the Future of Digital Identity in Africa

Biometrics promise a simple future: your face, fingerprint, or voice becomes your key. No passwords. Less fraud. Faster onboarding. Wider access. It is an attractive proposition in any market. In Africa, it is especially compelling because identity systems are often fragmented and unevenly trusted. Millions of people still struggle to access banking, healthcare, education, and government services because they cannot prove who they are reliably. Biometrics can help close that gap.

But there is a problem that rarely gets the attention it deserves. The same system that can unlock inclusion can also harden control. When biometric identity becomes the default gatekeeper, the discussion stops being about convenience and starts being about power.

This essay is not anti-biometric. It is anti-naivety.

Why biometrics are expanding so fast

Biometric adoption is not irrational. Three forces are pushing it forward aggressively.

1) Fraud pressure is relentless

Fintechs, telcos, and public services face constant identity fraud attempts: synthetic identities, impersonation, account takeovers, and document forgery. When fraud economics favour the attacker, businesses look for stronger assurance mechanisms. Biometrics are marketed as that stronger mechanism.

2) Convenience economics is real

Every extra step in onboarding reduces conversion. Faster onboarding improves growth metrics and reduces operational cost. Biometrics appear to offer speed without sacrificing assurance, so they become a product and revenue lever, not just a security control.

3) Policy and state capacity goals

Governments want single identity rails to deliver services efficiently, reduce leakages, and improve targeting. When public systems are under pressure, biometric identity becomes a tempting “one system to rule them all.”

None of this is surprising. The risk comes from what happens when biometrics are deployed faster than the safeguards required to keep them legitimate.

The central risk: biometrics are not resettable

Passwords can be changed. Tokens can be rotated. Cards can be reissued. Biometrics cannot. Not in any meaningful way.

If a biometric template is compromised, the user cannot simply “get a new fingerprint.” That means biometric overreach is not a normal data protection issue. It is a lifetime identity issue.

So the correct governance question is not, “Is biometric data secure?” That is the wrong framing because nothing is perfectly secure. The correct question is:

What happens when it is not secure, and can citizens recover without permanent harm?

If the answer is unclear, adoption at scale becomes irresponsible.

The second risk: function creep (efficiency today, coercion tomorrow)

Function creep is when a system built for one purpose quietly expands into others.

A biometric system begins as:

welfare distribution integrity

Then becomes:

SIM registration requirement

Then becomes:

banking and payments onboarding

Then becomes:

travel, border controls, and work access

Then becomes:

the default requirement for basic participation in public life

Each step can be justified as efficiency. Taken together, it can become coercion.

The boundary between the two is governance: clear purpose limitation, legal constraints, independent oversight, and meaningful penalties when boundaries are crossed. Without those, function creep is not an accident. It is the natural outcome of institutional incentives.

The third risk: exclusion by design

Biometric systems fail at the edges, and in African contexts those “edges” are not rare.

Common failure patterns include:

worn fingerprints for manual workers and artisans
variable lighting conditions affecting facial recognition
ageing and physical changes over time
disability and accessibility gaps
low-quality capture devices in field enrolment environments
inconsistent connectivity and synchronisation failures

If your identity system assumes a clean biometric match is always possible, you are not building inclusion. You are building a system that will quietly lock out vulnerable populations and then blame “failed verification.”

Inclusion cannot be a slogan. Inclusion must be engineered:

with fallback paths
with appeal mechanisms
with a clear standard for resolving edge cases quickly and humanely

A biometric-first system without fallback is not identity infrastructure. It is a brittle gate.

The fourth risk: surveillance normalisation

This is the broadest and most culturally significant risk. When biometric checks become routine, constant verification becomes normal. Over time, populations adjust to being measured, tracked, and evaluated as a condition of participation.

This is not paranoia. It is a governance question:

who has access to biometric systems and logs
under what authority
under what oversight
with what retention limits
with what transparency
and with what appeal mechanisms for citizens

If these questions are not answered, biometric identity can become a tool of asymmetric power, even if it began with good intentions.

Biometrics do not remove fraud, they shift fraud

There is another reality leaders often ignore. Biometrics do not eliminate fraud. They change the fraud economy.

When attackers cannot easily impersonate a user at login, they move to:

enrolment fraud (getting enrolled as someone else)
capture fraud (manipulating biometric capture in weak field conditions)
insider collusion (staff facilitating fraudulent enrolment)
account recovery abuse (resetting or bypassing biometric requirements)
device and session hijacking (taking over a “verified” session rather than identity)

A biometric programme that ignores these pathways will overinvest in enrolment while leaving the operational attack surfaces open.

What responsible biometric deployment looks like

A serious biometric strategy must include safeguards that are operational, auditable, and enforceable. Anything less becomes a high-risk social experiment.

1) Privacy by design and data minimisation

collect only what is necessary for a specific purpose
store as little as possible, retain for the shortest time
avoid centralised storage unless truly required
deletion must be real, verifiable, and enforceable

Minimisation is not cosmetic. It reduces blast radius.

2) Strong cryptography and template protection

protect biometric templates with modern cryptographic methods
harden central stores to extreme levels if centralisation is used
enforce strict access controls and logging
conduct independent security audits and publish high-level results

The security posture should match the irreversibility of compromise.

3) Independent oversight and transparency

Internal governance is insufficient. For national-scale biometric systems, credibility requires:

independent oversight mechanisms
published governance frameworks
clear breach handling protocols
transparency reports that show how data is accessed, by whom, and why

Trust collapses when the public suspects silent misuse.

4) Human fallback, alternatives, and appeals

No biometric system should be a single point of exclusion. There must be:

alternative verification paths for edge cases
time-bound escalation and appeals
mechanisms to correct identity errors
clear service standards so people are not trapped in bureaucratic loops

A system that cannot handle exceptions is not inclusive. It is punitive.

5) Purpose limitation with meaningful penalties

Function creep is inevitable without consequences. Policies must include:

strict purpose limitations
audit rights
penalties for misuse
constraints against secondary use without renewed consent and authorisation

If penalties are weak, boundaries will be crossed.

6) Operational controls against insider collusion

Biometric programmes must explicitly address insider risk:

separation of duties for enrolment and approvals
anomaly detection for unusual enrolment patterns
random audits and sampling of enrolment quality
strong discipline for violations

Otherwise, the “trusted enrolment process” becomes the fraud point.

The business case for restraint

Some leaders assume privacy and safeguards slow growth. That is short-term thinking.

Biometric trust failures produce:

reputational collapse that spreads faster than technical remediation
regulatory backlash and forced redesign
user abandonment and increased reliance on informal channels
long-term distrust in digital identity programmes

Restraint is not weakness. Restraint is durability.

The organisations that win will be those that treat biometric identity as critical infrastructure with governance, evidence, and redress, not as a growth hack.

The future needs guardrails

Biometrics can be a foundation for inclusion in Africa. They can also become a foundation for irreversible harm if deployed recklessly.

The principle is simple: identity systems must protect dignity, not just reduce fraud.

Africa’s digital identity future will not be judged by how many people were enrolled. It will be judged by how many people remained protected, included, and free from silent coercion.

If we want biometric identity to become infrastructure, it must be governed like infrastructure: designed, audited, monitored, and constrained by rules that outlast whoever is in power today.