The cost of digital ID verification may seem like a technical detail, but it lies at the heart of expanding access to financial services, supporting innovations such as open finance, and enabling a digital future. In June 2025, South Africa’s (SA’s) Department of Home Affairs (DHA) announced that it would increase the price of real-time verifications against the National Population Register.

The price of digital ID verification, which underpins Electronic Know Your Customer (eKYC) authentication in the banking sector, rose from just R0.15 to R10 in peak hours, and to R1 for off-peak batch verifications. This 6,500% increase has sparked debate across the financial sector – and in our own offices at Cenfri – about whether the new price is fair: fair for whom, and fair in terms of which outcomes?

During the debate, two differing points of view emerged at Cenfri – an infrastructure perspective championed by Barry Cooper, and an open-finance perspective advocated by Jeremy Gray. Their views have been set out here in dialogue form to give you a glimpse into the kinds of conversations that shape our thinking at Cenfri.

Is the price fair? Around 180,000 digital ID verifications are processed daily in SA. At R1 (~USD0.053) per batch check, this would generate R65.7m per year; at R10 per real-time check, closer to R657m. For context, SA’s new rate of R10 (~USD0.53) is more than double the global average (~USD0.20) and five times the regional average (~USD0.11).

Jeremy:

The question of fair pricing for ID verification mirrors wider challenges we’ve seen in our work across Africa: how to make open finance, the mandated sharing of consumer data like identity information between financial service providers (FSPs), based on consent, both technically feasible and financially sustainable. Open finance requires every FSP to develop open APIs so that any other FSP can request consumer data, provided they meet security requirements and have consent.

This means that on a technological level, each FSP must provide a service similar to what the DHA offers. However, these FSPs would need to share and verify data on multiple use cases, compared to the DHA that would only be sharing data for one use case (ID verification).

Brazil, considered the global leader in open finance, processes more than 450 million API calls per day, and the cost of their implementation since 2020 is about USD44m, or R770m, spread over four to five years. This is about the same as what SA’s DHA will earn per year processing 180,000 API calls per day in just one year. The question, therefore, is whether the development and maintenance of a reliable open API for real time ID verification really costs R657m per year. Only the DHA can answer this honestly.

Barry:

Jeremy raises the issue of cost in comparison to global examples, but from an infrastructure perspective, there is a broader question. Many developing countries struggle to leverage digital ID infrastructure at a price that supports not only financial inclusion but also deeper, more sustainable participation in the economy. In India, even with a population of 1.45 billion (2024) driving 430 million ID authentications per month, it still costs institutions around R3.88 per successful real-time authentication. Comparing the Indian scale against the 5.5 million per month in SA, means that similar functional national infrastructure is driven on 1.27% of the Indian volume.

The experience from our work across African countries points to scale issues; small populations with larger relative fixed-cost components for population register infrastructure and networks. Limited scale results in use cases not being affordable for consumers and hence not directly recoverable. Government-funded development infrastructure expenditure would be a more reliable way to achieve scale within a viable timeframe. On top of this are the enrolment and ongoing operational costs, which need to be covered by the activities they support.

So, what could you actually get for your R10? Firstly, the R10 price is clearly intended to divert verification traffic from the real-time process to overnight batch processing to take pressure off the real-time channel and improve overall response rates. In the South African case, you are promised quite a lot – top-rate biometric verification and biographic identity checks at vastly improved uptime. If, however, uptime continues to hover at 50% and citizens with [physical] green ID books are excluded, then R10 is not a good deal.

The DHA will not receive a massive boost in fees based on a simple straight-line projection using the new price and historic traffic, but will more likely receive a modest positive bump in fees, which in theory could help support improved infrastructure to fix longer-term systemic failures. A cheap service that doesn’t work for all ID holders, nor fails to achieve basic uptime targets for the financial sector, isn’t really worth that cheap price.

Who should bear the cost? Although no single party will bear the entire cost, the government has shifted much of the responsibility onto the private sector. This raises the issue of who should shoulder the burden, and whether that burden is fairly shared.

Barry:

An aggravating factor is that in some jurisdictions, particularly state-owned and state-run facilities, inefficiency is rife. Who should pay for these inefficiencies is a thorny question. The private sector, though often more efficient, has inherent mandate and incentive problems, while the public sector has the mandate but often lacks incentives and cohesive direction. Banks and fintechs can’t bear both public inefficiencies and higher fees. Ultimately it will be the consumer who pays for inefficiencies through either higher financial services fees or reduced financial sector resilience.

Perhaps the state should subsidise or rethink models along the lines of a public-private partnership where the state brings the mandate and the private sector competes on efficiency and innovation?

Jeremy:

I agree, seamless ID verification is a public good, but increasing its cost of verification also raises the cost of all personal data sharing, with knock-on effects not just for financial services but also for open finance and the wider digital economy. One look at the earnings reports of the large financial institutions shows that the largest banks can clearly absorb these costs. The concern is that this will have a far more harmful effect on smaller innovators and competitors, and the large banks are almost certain to simply pass these costs onto the consumers. The net effect in both cases is bad for the consumer. Other countries, such as India, Estonia and Rwanda, provide it free or at low cost.

What are some other considerations? One of SA’s national goals is financial inclusion, alongside the broader ambitions of digital transformation and delivering public value through government services. The question, therefore, is not only whether R10 covers the costs of ID verification, but whether it helps or hinders these outcomes.

Jeremy:

Increasing the cost of ID verification increases the costs of all personal data sharing. This not only affects existing financial services, as highlighted by Tyme Bank, but has even more substantive implications for the implementation of open finance in SA — an intervention that could deliver more tailored products at lower cost. It also has broader implications for society.

The Presidency and Treasury are spearheading efforts to digitise government services through the MyMzansi initiative, most of which will rely on regular ID verification. SA may not be in a financial position to provide verification free of charge, as in India, Estonia and Rwanda, but the increase appears at odds with core policy priorities, hindering both private-sector innovation and citizen outcomes.

Barry:

On a good day, with system uptime stuck at 50% and access limited to those with ID cards, we fall far short of true financial inclusion and remain a long way from unlocking real digital transformation. The core problem with the previous pricing and service model was that it was not sustainable, causing financial institutions to skip building internal digital ID registries and instead polling the national registry. They were verifying the same static credentials multiple times per client per year or month, despite most of those credentials remaining constant for life, essentially verifying credentials that were already verified, instead of validating the consumer against the verified credentials already on their system.

At a low price there is always endless demand. The alternative route that the new pricing regime seeks to increase is the overnight batch verification at R1, which has a much higher uptime. Customer-due-diligence does not need to be instant; a 12 to 24 hour delay can easily be accommodated in most circumstances, with sufficient national ID validation checks as well as risk-based product design leveraging the Financial Intelligence Centre Act and FATF guidance instead of defaulting to expensive real-time verification.

Verifying already verified identifiers is just wasteful compliance and indicates a limited grasp of risk and a risk-based approach to CDD. Real-time verification should ideally be for higher-risk accounts, where the account is required to be used almost immediately, a big red flag in any event.

So, what has really changed is that the verification price has increased to R1 (~USD0.053) and the service level has changed to 12 to 24 hours. The DHA is no longer an outsourced customer relationship management (CRM) registry for financial institutions. In the end, fairness must be judged not only by the price tag, but by whether the system effectively expands access, supports inclusion, and delivers real public value.

Take home points: It is clear that there is a need to improve the DOH identity verification system however, who should bear the cost of this upgrade and whether the cost is justified is not a straightforward answer. Compared to global benchmarks, R10 looks excessive unless system reliability and coverage improve. When it comes to who should pay, neither the private sector nor government can carry the full burden; a balanced approach is needed. And in terms of outcomes, higher fees risk undermining SA’s goals of financial inclusion, digital transformation and public good. In the end, whether the cost increase is justified will be measured not in rands and cents, but in how many citizens are included in the digital future.