Building the Confidence Architecture

The CBN and NIMC should jointly establish a binding interoperability standard for identity verification across financial institutions, beginning with systemically significant institutions and supported by shared infrastructure investment and multi-stakeholder governance.

Progress on BVN-NIN linkage is meaningful, but interoperability remains incomplete. Integration costs and system reliability continue to constrain scaling, while fragmented verification processes weaken fraud prevention, accountability, and consumer recourse.

This would require treating identity infrastructure as a shared public good rather than as a fragmented institutional compliance requirement, with corresponding implications for funding, governance, and privacy design.

The CBN should require systemically significant financial institutions to participate in a mandatory fraud intelligence-sharing platform, governed by an independent body with regulatory oversight and supported by privacy-preserving protocols that address competitive sensitivity while creating binding collective defence obligations.

The evidence for coordination is already on the record: under NeFF, fraud losses fell 51% to ₦25.85 billion in 2025, and a Person of Interest Portal now tracks more than 13,400 individuals linked to fraudulent activity. But NeFF remains advisory and voluntary, fraud reporting fell 34% in the last quarter of 2025, and fraudsters continue to move across institutions while reporting remains inconsistent.

This challenge cannot be resolved through voluntary cooperation alone. Effective solutions must align private incentives with collective benefit, making participation in shared defence mechanisms rational for individual institutions as well as mandatory where the system requires it.

Financial sector regulators should publish binding AI governance standards for financial services, requiring board-level oversight of model deployment, regular bias auditing and performance monitoring, explainability standards for automated decisions in credit, fraud, and customer management, and consumer contestability mechanisms for algorithmic outcomes.

The CBN sandbox, the National AI Strategy launched in April 2025, and the AI-powered AML guidelines issued in March 2026 provide meaningful foundations, but they address AML compliance specifically rather than the full range of AI applications in financial services. Practitioners located the greater danger not in the technology itself, but in the governance failures surrounding its deployment.

These standards should be developed in consultation with industry and international regulatory peers so they shape practice early, before fragmented deployment creates larger confidence shocks.

Financial sector regulators should commit to a phased transition toward outcome-based supervision, beginning with a defined set of digital financial services and accompanied by standing industry-regulator dialogue mechanisms with clear mandates and regular senior-level participation.

Current engagement models remain heavily process-driven and are not always well matched to the system-level risks emerging in platform-led finance. Participants pointed out the need for more substantive dialogue on multi-party accountability, emerging technologies, and systemic interdependence.

This transition should be accompanied by structured engagement mechanisms that support earlier problem-solving, clearer feedback, and more effective coordination across the system.

Financial institutions should treat dispute resolution as strategic infrastructure rather than as a cost centre, extending resolution coverage across all active channels, establishing clear timelines and accountability frameworks, developing consumer-facing tracking tools, and measuring resolution quality as a board-level performance indicator.

The CBN’s Industry Dispute Resolution System (IDRS) for card channels and its draft 48-hour refund rule are important steps. But consumer complaints surged 143% in the first half of 2025, and the IDRS covers card-based channels only; mobile, USSD, and platform-based transactions still lack equivalent coverage. Resolution speed, not failure prevention alone, emerged as a primary determinant of consumer confidence.

Investment in recourse should therefore be treated not as an operational burden, but as a structural determinant of whether users remain in the system.

Financial institutions should establish board-level AI governance frameworks before regulation compels them to do so, including independent audits of model performance, governance arrangements for bias monitoring and explainability, and clear senior accountability for AI oversight.

Institutions should not wait for regulatory requirements to establish governance discipline. The gap between AI deployment and AI governance already represents a material risk.

Institutions that move early will be better positioned not only to manage risk, but to shape the standards that follow.

Industry associations and financial institutions should build the neutral intermediary capacity required for collaboration on trust across the ecosystem. This should include a standing governance body that brings together banks, fintechs, mobile money operators, and regulators around shared threats and systemic standards.

This goes beyond fraud intelligence sharing alone. It requires building a cross-cutting coordination mechanism that neither NeFF nor open banking currently provides. Participants identified institutional mistrust as the single greatest barrier to collective action, which means the problem is not only technical design but relationship architecture, extending beyond the reach of regulation to require active industry leadership.

Development partners should support the design and sustained funding of a longitudinal consumer trust survey covering Nigeria’s financial system, with comparable methodology across frontier market economies.

The Edelman Trust Barometer covers Nigeria only at an economy-wide level, and IPA’s consumer protection surveys, while valuable, are not a recurring sector-specific measure. The case for such an instrument is now reinforced by policy: the Central Bank’s Payments System Vision 2028 commits to launching a National Payments Trust Index, with a target score of 80 by 2028 and a methodology to be developed in its early phases. That commitment establishes, at the level of national strategy, that trust in the payments system is a variable to be measured rather than assumed. The instrument itself, however, does not yet exist, and its value will depend on how it is built.

A demand-side barometer, designed from the outset to be comparable over time and across markets and to capture confidence as consumers actually experience it, is the instrument this domain requires. Its credibility will rest on design choices made early: measuring what users experience rather than what providers report, independent computation, comparability across markets and over time, and resistance to becoming a target that is managed rather than a condition that is earned.

International development institutions should fund the development of shared, audited AI training datasets relevant to African financial systems.

No shared federated infrastructure currently exists for AI model training in fraud detection or credit scoring in Nigeria. Institutions develop proprietary models on proprietary datasets, producing inconsistent model quality, duplicating costs across the sector, and weakening collective accountability.

Shared datasets would improve model accuracy, reduce individual institutional cost burdens, and create a mechanism for embedding fairness standards into the training data shaping AI deployment across the ecosystem.

Development partners should support mechanisms for cross-country learning that enable frontier market policymakers and practitioners to share evidence and adapt governance innovations across contexts.

Nigeria’s removal from the FATF grey list in October 2025 and the CBN Fintech Report’s proposals for bilateral regulatory passporting with Ghana, Kenya, and Senegal both signal a strengthening of the country’s standing within global financial governance frameworks. The structural challenges identified at the Bridgforte Executive Table – around identity, accountability, AI governance, and regulatory design – are not unique to Nigeria.

The opportunity is to translate that standing into intellectual and policy leadership, positioning Nigeria not only as an adopter of global standards, but as an exporter of governance frameworks.

Recommendation	Industry-wide Fraud Intelligence (70%)	Ethical AI & Data Governance (65%)	Shared Consumer Protection (55%)	Structured Regulatory Dialogue (50%)
R1 Mandate Identity Infrastructure
R2 Mandatory Fraud Intelligence Sharing
R3 AI Governance Standards
R4 Outcome-Based Supervision
R5 Dispute Resolution Investment
R6 AI Governance for Institutions
R7 Cross-Sector Collaboration
R8 Longitudinal Trust Barometer
R9 Shared AI Datasets
R10 Global Cross-Learning

Related in this report