Navigating BSP AI requirements: A Philippine lenders’ roadmap

The Bangko Sentral ng Pilipinas (BSP) has made artificial intelligence governance a cornerstone of its regulatory framework. For Philippine financial institutions navigating this evolving landscape, understanding these guidelines isn't optional, it's foundational to remaining compliant and competitive.

As the central bank continues to refine its approach to AI adoption in banking, lenders face a critical question: how do we harness AI’s transformative potential whilst meeting stringent regulatory expectations? 

The BSP’s AI Governance Framework 

The BSP recognises that AI is reshaping financial services. Unlike a one-size-fits-all approach, the central bank has developed guidance that acknowledges AI’s benefits whilst addressing systemic risks. The framework emphasises transparency, accountability, and robust governance structures. 

Key pillars of the BSP’s guidance include: 

Risk management and oversight 

Financial institutions must establish clear governance frameworks for AI systems, with defined roles, responsibilities, and escalation procedures. This means moving beyond treating AI as an IT project—it requires board-level engagement and strategic oversight. 

Data integrity and security

As AI systems rely on historical data to make lending decisions, customer onboarding recommendations, and fraud detection, data quality becomes non-negotiable. The BSP expects institutions to maintain rigorous standards for data governance, validation, and protection, often aligned with the Data Privacy Act of 2012. 

Fairness and bias mitigation

AI systems can inadvertently perpetuate lending bias. The BSP’s guidance requires institutions to audit algorithms regularly, document their decision-making processes, and demonstrate that AI-driven outcomes do not discriminate against protected groups. 

Transparency and explainability

When an AI system denies credit or flags a transaction, customers and regulators must understand why. The BSP expects financial institutions to maintain explainable AI systems, not black boxes. 

Why this matters for Philippine lenders 

The Philippines has over 100 million adults, with roughly 70% lacking access to formal banking services. For microfinance institutions, rural banks, and expanding commercial lenders, AI presents an unprecedented opportunity to reach underbanked populations through faster credit assessment and lower operational costs. 

However, this opportunity comes with regulatory guardrails. Consider the trajectory: as the BSP continues to strengthen cybersecurity requirements and digital financial inclusion targets, institutions that embed governance from day one gain a strategic advantage. 

Rajan Uttamchandani, CEO of Esquire Financing, highlighted the importance of partnership in this journey: “In an era of the everchanging technological landscape and higher demands from clients, it is paramount to have a trusted partner who delivers on technology and supports you navigate from a business perspective.” 

Building AI capability within regulatory constraints 

Many Philippine lenders worry that strict AI governance will slow innovation. The opposite is often true. Institutions with robust AI governance frameworks move faster because they can scale with confidence. 

Here’s why: 

  • Operational clarity: When your institution has documented AI governance, every team member understands their role. Lending officers know which decisions are algorithmic and which are human-led. 
  • Faster regulatory engagement: When the BSP has questions about your AI deployment, a well-documented governance framework allows you to respond swiftly. This reduces implementation timelines and regulatory friction. 
  • Customer trust: As fraud remains a concern in Philippine financial services, institutions that transparently explain their AI decisions build customer confidence. This translates to lower churn and stronger market positioning. 

George Chesakov, CEO of Salmon, spoke to this advantage: “Oradian offered us both the beauty of being a cloud-hosted, ready-made solution, but with our ability to develop it further using modern software development techniques. It makes it so much more future-proof in terms of how you build your IT organisation.” 

Key areas of focus for Philippine institutions 

Credit Risk Assessment 

The BSP expects lenders to validate any AI-driven credit scoring system. This means historical back testing, forward-looking stress tests, and regular audits of approval rates across demographic groups. For rural and microfinance lenders, this is particularly important as your underwriting models must demonstrate that AI improves credit quality whilst maintaining financial inclusion objectives. 

Anti-Money Laundering (AML) and Sanctions Screening 

AI powers much of modern AML compliance. However, the BSP requires that AML AI systems maintain explainability. When your system flags a customer or transaction, investigators must understand the reasoning. This is non-negotiable for BSP-regulated entities under Circular No. 706. 

Customer Due Diligence (CDD) 

Know-Your-Customer (KYC) and Customer Due Diligence processes are increasingly AI-enhanced in the Philippines. The BSP expects these systems to be auditable, with clear documentation of how identity verification and beneficial ownership determination are conducted. 

Practical steps: The implementation roadmap 

Month 1-2: Governance and assessment 

Establish an AI governance committee with representation from lending, risk, compliance, and IT. Conduct a baseline assessment of existing use cases. 

Month 3-4: Documentation and transparency 

Document your AI decision-making processes. For credit systems, create “model cards” that explain inputs, outputs, and known limitations. 

Month 5-6: Testing and validation 

Back test AI models against historical data and conduct bias testing across protected groups. 

Month 7+: Monitoring and iteration 

Implement ongoing monitoring systems that track model drift. 

The role of technology partners 

Your core banking platform provider should be more than a vendor—they should be a governance partner. Your partner should provide: 

  • Pre-built compliance frameworks: Governance templates aligned to BSP expectations. 
  • Audit-ready architecture: Immutable audit trails that support regulatory reporting. 
  • Transparency tools: Features that allow you to demonstrate how AI decisions are made. 

Looking ahead: AI and Financial Inclusion 

The BSP’s AI guidance reflects a broader commitment: technology should expand financial access, not restrict it. According to the BSP’s National Strategy for Financial Inclusion, technology-enabled delivery channels are essential to reaching the country’s unbanked. 

Get the whitepaper: The digital-first bank’s guide to AI in 2026

The BSP’s AI guidance is not a constraint; it’s a roadmap. Institutions that act now gain first-mover advantage. Download our comprehensive white paper to get started: Digital Banking Guide: AI and Artificial Intelligence in 2026

Related Insights

Discover more insights

AI fraud detection guide for digital-first banks in 2026

Fraud in emerging markets just changed. Rule-based fraud detection systems are obsolete. Here's why: more than 50% of fraud now involves AI, but most detection systems can't adapt faster than fraudsters iterate.

The good news is that institutions using AI fraud detection can achieve 90-99% accuracy. But only if they solve data infrastructure first.

This playbook covers everything from deepfake-enabled account opening to fraud-as-a-service marketplaces. But the centrepiece is this: the same institutions that detect fraud in real-time are the ones whose cores process transactions as events, expose customer data through APIs, and let teams investigate without a vendor ticket.

Your fraud tools are only as good as your data layer. This guide shows you how to build it.

The regulation playbook for digital-first banks

Regulators in Nigeria, the Philippines, and Indonesia are asking one question: what can your infrastructure demonstrate in real time?

Here's the problem: most digital banks fail at this. They're still pulling compliance data from silos, reconstructing audit trails manually, waiting weeks to produce evidence regulators expect in minutes.

Before you hire another compliance officer or buy another AML tool, ask yourself: Can your team access a unified customer view across all channels in seconds? Can you explain why your AI model made a decision? Do you have complete audit trails, or are you stitching together spreadsheets?

This playbook shows you how to build the infrastructure that actually works. Unified data layers. Real-time monitoring. AI governance. But most importantly, why your core banking system matters more than any process, and how to fix it before your next compliance deadlines.

Think bigger. Go further.

Come and see the future with us. Talk to one of our core banking experts.