Dickie Shearer: Reframing How AI and Banking Are Being Built for the Global South
Qatar, December 20, 2025 - Dickie Shearer has become a key voice for AI-native financial infrastructure across emerging markets. His lens is simple and direct: build systems that reflect how local economies actually work, or accept persistent underperformance and exclusion.
He founded Tintra, which is set to launch AI-native financial technology and banking in 2026 across multiple markets in the Global South. His work centers on cultural intelligence, financial sovereignty, and a move toward a more multipolar financial order.
The core thesis: it's not access - it's cultural fit
Shearer argues that most global financial systems fail outside Western economies because they're built on assumptions that don't match local reality. Digitising those models just reproduces the mismatch faster.
"The issue has never been access," he has said. "It's cultural nuance. New problems require new solutions, and challenges in the Global South cannot be solved using frameworks designed for Western economies decades ago."
AI is not culturally neutral
AI inherits the data, priorities, and blind spots of its training environment. Systems trained on assumptions of formal employment, predictable cash flow, and standardized credit histories struggle where value moves informally, seasonally, or through trust-based networks.
Shearer's approach puts cultural intelligence at the center. Don't bolt AI onto legacy products. Build banking infrastructure that learns from local behavior and adapts policy, risk, and compliance to how value actually moves.
What Tintra is building
Tintra is developing AI-native cross-border banking and settlement systems for emerging markets. The architecture focuses on cultural intelligence, regulatory interoperability, and South-South trade.
The goal is operation across currencies and rule sets while staying grounded in local trade patterns and compliance realities. AI serves as an enabling layer that observes, learns, and adapts - not a blunt instrument chasing Western templates.
Why this matters for banks and insurers
In high-income markets, most economic life is visible to formal systems - bank accounts, credit rails, insurance, and tax networks. In many emerging markets, only a fraction is captured; the rest moves through informal trade, savings circles, and trust networks.
That visibility gap makes national economies look smaller than they are. Infrastructure can either constrain or amplify: design that reflects real behavior reduces friction, brings activity into view, and lets potential scale.
Practical steps for practitioners
- Data strategy: Blend alternative data (mobile money histories, utility payments, co-op and supply-chain records, remittance patterns) with consent-first collection and clear data minimization. Localize feature stores by market to avoid one-size-fits-all signals.
- Credit and risk: Move beyond bureau scores. Model event-based cash flows, seasonality, group lending dynamics, and trust-network guarantees. Use segmented performance monitoring and cohort-level fairness checks.
- Compliance operations: Encode policy-as-code with dynamic KYC tiers tied to risk and use-case. Weight transaction monitoring by network trust, and support local scripts, aliases, and transliteration in sanctions/name screening. Maintain corridor-specific rule libraries.
- Payments and settlement: Interoperate with domestic instant payment systems and mobile money. Map cleanly to ISO 20022, and support regional settlement options that reduce dependence on external intermediaries where regulation allows.
- Identity and verification: Support local ID schemes and community attestations where formal IDs are thin. Combine biometric fallbacks with clear redress paths.
- Model governance: Stand up local oversight with market experts. Stress-test for concept drift across seasons and harvest cycles. Keep humans in the loop for edge cases and adverse actions.
- Insurance design: Use non-traditional signals for pricing (weather, logistics reliability, cooperative history). Expand parametric covers for climate, crop, and outage risks. Make claims mobile-first with fraud checks tuned to local patterns.
- Treasury: Plan for local currency liquidity, corridor-level netting, and contingency routes. Monitor policy changes impacting FX, capital controls, and settlement windows.
Metrics that actually reflect progress
- Visibility: Share of economic activity captured from outside the formal sector; conversion rates from informal to semi-formal use.
- Risk quality: Default and loss metrics segmented by corridor, season, and network type; model stability across cohorts.
- Compliance efficacy: False positive/negative rates by customer segment; time-to-clear alerts; KYC completion time and model explainability rates.
- Settlement performance: Share of flows settled regionally; FX slippage and cost per cross-border transaction by route.
Risks to avoid
Applying Western assumptions to non-Western contexts can exclude the majority of users who move value informally. AI can make that exclusion worse by optimizing the wrong target.
As Shearer warns: "The danger is building ever more sophisticated intelligence on foundations that don't reflect reality." Start with cultural fit, then optimize.
Where policy and industry are heading
As more governments explore options beyond dollar-centric settlement and externally imposed models, sovereignty and resilience are taking priority. Banks and insurers that adapt early will be better placed to serve South-South flows and cross-border trade grounded in local behavior.
For wider context on cross-border modernization, see the BIS work on improving cross-border payments here, and data on financial inclusion dynamics via the World Bank's Global Findex here.
Bottom line
Stop trying to squeeze emerging markets into legacy blueprints. Build AI-native infrastructure that learns from local behavior, respects sovereignty, and connects regional trade with compliant rails.
If your team is standing up AI capability for financial use cases, this curated set may help: AI tools for finance. It's a practical way to spot gaps in modeling, data, and workflow automation before they scale.
Your membership also unlocks:
