Digital public infrastructure is scaling faster than the means to correct it when it fails. Our new report led by David Porteus, founder of Integral: Governance Solutions and co-authored by Sahana Arun Kumar together with Deepti George, Founder of Yutadhi, and Maria Luciano, Tech Policy Researcher, looks at what that means for the people using it.

The Universal DPI Safeguards Framework calls for “effective remedy and redress” but does not define “effective” or say how to measure it. With more than 39 countries committed to launching at least one DPI component by 2028, that gap matters. When a stolen credential shows up as a failed payment, or outdated registry data leads to a denied benefit, people need somewhere to go, and weak redress erodes the trust that adoption depends on.

Between May and July 2026, teams in Brazil, India, Benin, Ethiopia and South Africa studied 15 systems: one instant payment, one digital ID and one data-exchange system in each country. Our teams led the fieldwork in Benin, Ethiopia and South Africa, using document review, mystery shopping of grievance channels and interviews with operators and key informants.

What we found:

→ Grievance redress was complex, opaque and fragmented. Users mostly deal with participating banks or providers, not the scheme operator, so operators can lose sight of how first-tier providers perform.

→ Only 2 of the 15 systems regularly publish GRM volumes. Pix and UPI both do, but they count different things (consumer-initiated cases versus bank-originated chargebacks), so the numbers can’t be compared. More disclosure doesn’t create comparability without common definitions.

→ On liveness, a minimum threshold based on public information, 4 of the 15 GRMs were fully live, 8 were underway and 3 were not live.

The report proposes two practical tools. Liveness asks whether a functioning GRM is visibly in place. Wellness is a scorecard of how well it performs, which needs operator data and could be completed in about 20 working days where that data exists. It also sets out the case for a common grievance taxonomy and data object, so a complaint can move between providers, schemes and DPIs without the complainant starting over, with privacy limits on collection, authentication and retention built in from the start.

The work is exploratory: five countries, 15 systems, and measures that still need piloting. But the direction is clear. Infrastructure that cannot correct its own failures, or that is not designed with redress from the start, will not earn the public’s trust.

Recognising efforts of Amarante’s country experts in Africa who made the field work possible without which this publication would not have been possible: Girum Fekadu Diriba (Ethiopia), Antonia Bassène (Benin) and David Nsenginuyama (South Africa), and to the operators and informants who gave us their time.

 Read the full report: https://www.integralsolutionists.com/GRMINDPI