The Complete Guide to Using AI in the Financial Services Industry in Peru in 2025

Peru's financial sector is at an inflection point: regulators have opened the door to

100% digital banks

and pushed payments interoperability, while a new national AI regime raises the stakes for safe, explainable systems - so getting AI right is now a business and compliance priority.

Local law and practice show AI already powering fraud detection, real‑time AML/CFT monitoring, chatbots, alternative credit scoring and regtech automation across banks and fintechs (see the Fintech 2025 market overview), and Peru's Law 31814 frames AI use with a risk‑based approach, human oversight and data governance obligations.

That combination - rapid fintech growth, multiple supervisors (SBS, BCRP, SMV, UIF‑Perú, INDECOPI, APDP) and clearer AI rules - means Peruvian firms that pair practical pilots with documented risk controls win trust, reduce losses and unlock faster, fairer access to credit for more consumers (Fintech 2025 Peru market overview, Peru Law 31814 AI regulation overview).

Peru's AI rulebook is no afterthought: Law 31814, enacted to promote AI for economic and social development, sets a clear, risk‑based framework that separates prohibited uses (like certain mass surveillance and manipulative systems) from high‑risk applications such as biometric ID, credit scoring and health or education tools, while insisting on human oversight, explainability and tighter data governance (Peru Law 31814 risk‑based AI framework details).

Oversight is centralized under the Presidency of the Council of Ministers via the Secretariat of Government and Digital Transformation (PCM‑SGTD), which will coordinate with data protection and sectoral bodies and run practical measures - think a national AI sandbox, public cloud for public‑interest projects and the CNIDIA innovation hub - to balance innovation and safeguards; the implementing Regulations (Supreme Decree No.

115‑2025‑PCM) published in September 2025 give private firms concrete duties on transparency, human review and documentation, including registration/traceability rules and minimum three‑year retention for impact records that create an auditable trail for supervisors (Regulations approving Peru Law 31814 (Supreme Decree No. 115‑2025‑PCM)).

For banks and fintechs the message is straightforward: classify systems by risk, build human‑in‑the‑loop controls, and treat compliance as an operational capability - because in Peru the regulator expects evidence, not promises, and timelines for sensitive sectors (finance, health, justice, education, security) start within a year of the rules coming into force.

Peru's fintech market in 2024–25 is a fast‑moving mix of scale‑up startups, greener regulatory soil and pragmatic AI pilots: regulators have cleared the way for 100% digital banks while SBS sandboxes and open‑payments rules are turbocharging real‑time interoperability, so wallets and neobanks can compete head‑on with incumbents (Fintech 2025 Peru market overview).

The numbers tell the story - the EY Peru FinTech Index counted about 237 active fintechs with roughly 17% year‑over‑year growth, and digital wallets now handle more than 347 million monthly transfer operations - a level of everyday use that explains why Yape alone has crossed more than 17 million users as firms push QR‑first payments into the provinces (EY Peru FinTech Index 2024).

AI is no longer experimental either: roughly a third of firms report using AI for fraud, alternative credit scoring and regtech automation, so successful projects tie clear business KPIs to documented governance and human‑in‑the‑loop controls that regulators now expect.

Peruvian banks and fintechs are already turning AI into tangible customer‑facing wins: the largest example is Banco de Crédito del Perú's One‑click cross‑sell, a cloud‑backed stack (BigQuery, Cloud Functions, Pub/Sub) that slashed the sales funnel to a single modal and drove dramatic results - online sales doubled in the first month, digital insurance conversions surged (a reported 15‑fold uplift in some metrics) and after two months One‑click made up over 70% of card insurance sales; read the full BCP case study on how cloud + propensity models fuel fast pilots (BCP One‑click cross‑sell on Google Cloud).

At the same time BCP is testing decisioning platforms to automate pricing and customer management, showing how rules + models speed time‑to‑market for offers (FICO decisioning for pricing and customer management).

These examples - simplified journeys, propensity scoring and automated decisioning - illustrate practical, regulator‑relevant pilots that other Peruvian firms can replicate at cloud speed.

“We managed to narrow down the information the user must provide to a single field from a modal without altering their browsing experience. Fewer steps in the purchase process significantly increase the conversion rate.” - Diego Nasra, Product Owner, Banco de Crédito del Perú (BCP)

Start every Peruvian AI project with a short, practical checklist that maps the system's risk class (Law 31814's prohibited / high‑risk / acceptable tiers) to concrete controls: (1) Classify the use case and run a Privacy/AI Impact Assessment - high‑risk uses (credit scoring, biometrics, employment, health) require full PIAs, human‑in‑the‑loop safeguards and documented explainability under the national AI regime (Peru Law 31814 risk-based framework); (2) Data protection basics: record and register personal databases with the ANPD, secure explicit informed consent (specially written consent for sensitive data), apply minimization and purpose limitation, and use contractual clauses or approved safeguards for cross‑border transfers consistent with the PDPL and its 2025 Regulation (Peruvian Data Protection Law (PDPL) and 2025 Regulation overview); (3) Security & ops: implement ISO/IEC 27001‑aligned technical and organizational measures, maintain a Security Document with access/privilege controls, backups and monitoring, and retain impact/traceability records for the multi‑year period regulators expect (audit trailability is non‑negotiable); (4) Governance & people: appoint a Personal Data Officer when thresholds apply, train model reviewers, and lock processor contracts so third parties cannot repurpose Peruvian data; and (5) Incident & audit playbook: prepare a 48‑hour notification workflow for large breaches, maintain continuous monitoring, and keep audit‑ready evidence that the AI lifecycle met safety, explainability and human‑oversight requirements.

Treat this checklist as operational - not legal wallpaper - and embed each control in the project timeline so audits, sandbox tests and supervisory questions are handled before launch (Peru data protection compliance roadmap).

Turn AI pilots into repeatable products by following a short, practical playbook that Peruvian banks and fintechs can actually execute: start every project with a defined business KPI and a staged pilot that proves value before widening scope, because roughly two‑thirds of financial AI projects stall before production and scale only after architecture and governance are fixed (Scaling Agentic AI: From PoC to Product in Financial Services); build a strong data foundation (domain embeddings, vector search and persistent memory for agents) so models plug cleanly into core systems; standardize on a unified platform and Responsible AI rules to lock in monitoring, explainability and cross‑functional workflows - only a small fraction of firms today have fully integrated standards, so treating standards as a feature accelerates ROI (2025 State of Responsible AI in Financial Services report); bake human‑in‑the‑loop checkpoints into decisioning for high‑risk flows, instrument traceability for audits, and measure lift with short A/B windows; and finally, learn from local practitioners and industry collaboration - Peru's leading banks and associations are already sharing lessons on algorithmic trading, fraud and compliance pilots, creating a practical playbook that others can adapt (GARP LATAM session on AI pilots in Peru (algorithmic trading, fraud, compliance)).

Follow these steps and a stalled proof‑of‑concept becomes a platform capability that delivers measured savings, faster product cycles and auditable, regulator‑ready evidence.

Risk management and cybersecurity for AI in Peru's financial sector must evolve from checklist thinking to an operational, real‑time discipline that blends model risk management with active cyber defenses: Peru's recent regulatory boom is a useful warning about performative rules unless firms build measurable controls (Harvard Kennedy School analysis: Peru's AI regulatory boom - quantity without depth), and financial institutions should treat AI both as a shield and a target - using machine learning to spot fraud, AML and anomaly patterns while defending models from manipulation and data drift (Corporate Compliance Insights: AI's dual role in financial services risk management).

Practical priorities are clear from model‑risk thinking: keep a current model inventory and audit trail, automate validation and monitoring so silent degradation is detected before it becomes a ticking time bomb, run generative‑AI stress scenarios and synthetic data tests, insist on explainability and independent validation for high‑risk flows, and lock third‑party contracts and oversight to avoid vendor blind spots.

Combine those governance steps with AI‑driven anomaly detection and behavioral biometrics in production to reduce fraud friction while preserving customer experience; the goal is not to ban complexity but to make it observable, auditable and cyber‑resilient so supervisors and customers can trust the results.

Peru's AI ecosystem is starting to attract practical, project‑level funding and cross‑border partnerships that financial institutions and universities can tap: for example, a targeted U.S. Embassy Lima Notice of Funding Opportunity - “Navigating Technological Risk in the Age of AI” - offers a $40,000 award to support a national conference that builds academic leadership on AI risks and responsible tech use (U.S. Embassy Lima NOFO: Navigating Technological Risk in the Age of AI), while the Embassy's Small Grants Competition (PD‑LIMA‑FY25) funds projects from roughly $10,000–$30,000 for seminars, exchanges and capacity‑building with a required U.S. component and SAM/UEI registration (U.S. Embassy Lima Small Grants Competition (PD‑LIMA‑FY25) funding details); at the regional level, the OAS‑backed initiative backed by over $1.1 million is building an AI policy framework for the Americas, creating partnership channels for Peruvian actors to influence standards and access skills programs (OAS-supported AI policy framework initiative for the Americas).

Practical next steps for banks and fintechs: identify eligible partners (universities, NGOs), start SAM/UEI registration early, and structure proposals around measurable capacity building or pilot governance work that complements Peru's domestic AI and financial rules.

Get started with a short, practical checklist: first, bookmark the Peru Government Contact List - ministry & regulator contacts so you can quickly find the right ministry or regulator (it even lists department, title, email and phone fields) and avoid delays when you need permits or sandbox sign‑offs; next, study local, replicable playbooks - like the Videsk–BCP contact-center video case study for Peruvian financial services - to see how cross‑border tools and cloud stacks actually streamline operations and customer journeys; and finally, invest a little time in team reskilling - target roles such as AI supervision and prompt engineering so staff can validate automated outputs and keep projects regulator‑ready with an upskilling guide for AI supervision and prompt engineering.

Together these three moves form a compact, local roadmap that turns curiosity into a documented, compliant pilot - think of it as packing a travel kit before you enter the regulator's sandbox.

Getting started with AI in Peru's financial services in 2025 means pairing fast, measurable pilots with regulator‑grade governance: pick one clear business KPI, run a short staged pilot, classify the system under Peru's Law 31814 and document human‑in‑the‑loop controls and impact records so supervisors can follow a three‑year audit trail; this is how firms move from experimentation to trusted production (see the national risk‑based AI framework at Peru Law 31814 AI risk-based framework).

The market is ready - digital banks, real‑time payments and growing AI use in fraud detection and credit scoring make 2025 a moment to act, not wait (Fintech 2025 Peru market overview).

Start small, instrument everything (model inventory, drift monitors, explainability docs), and train people to supervise models - practical, workplace‑focused courses like the AI Essentials for Work bootcamp can help teams learn prompt engineering, model oversight and operational AI skills before scaling (AI Essentials for Work registration).

In short: run governed pilots, prove measurable lift, keep inspectors able to “flip through” a clear audit trail, and you'll turn regulatory duty into a competitive advantage.