The BSP's draft circular on Guidelines on Model Risk Management, published as a policy exposure draft in June 2025, would require supervised institutions that use models to adopt a model risk management framework across the model life cycle. It defines a model to include methods that apply artificial intelligence and machine learning, and it covers vendor and third-party models with the same rigour as internal ones. Institutions would have two years from the circular's effectivity to comply. The BSP's June 2026 AI memorandum still described the framework as forthcoming.
OFFICIAL TEXT: bsp.gov.ph ↗ · PROPOSED · PHILIPPINES (BSP)
| Document | BSP draft circular on Model Risk Management — Draft Circular on the Proposed Guidelines on Model Risk Management |
| Issued by | Bangko Sentral ng Pilipinas (BSP), the central bank and supervisor of banks, non-bank financial institutions and payment system operators in the Philippines |
| Type | Consultation |
| Status | Proposed |
| Published | Jun 24, 2025 |
| Applies to | BSP-supervised financial institutions that use models, including non-bank credit card issuers to the extent possible. A policy exposure draft: not binding until issued as a circular |
| Official source | bsp.gov.ph ↗ |
| Use cases | Model risk management · Third-party & vendor AI · Credit scoring & underwriting · AI governance (general) |
What are the key points of BSP draft circular on Model Risk Management?
- Would amend Section 146/146-Q of the Manual of Regulations for Banks and for Non-Bank Financial Institutions on operational risk management to add management of model risk.
- A model is 'a quantitative method, system, or approach that applies statistical, economic, financial, or mathematical theories, techniques, and assumptions, including artificial intelligence and machine learning, to process input data into output.'
- Applies to all models regardless of origin, whether developed internally, acquired from vendors or embedded in third-party systems, proportionate to size, complexity, risk profile and extent of model use.
- The model risk policy must cover model identification and inventory, a tiering approach, standards for development, validation and monitoring, change approval, and management of model vendors.
- Vendor and external models must be integrated into the framework with the same rigour as internal models, with documented customisation and validation.
- Independent review of the framework, typically by internal audit or an external auditor, reporting to the board.
- Transitory provision: compliance within two years of effectivity, with an implementation program due within the first year.
What did BSP draft circular on Model Risk Management change for banks?
The draft would make model risk management, including for AI and machine learning models, a binding BSP requirement. The date is the earliest on which the draft is confirmed as posted on the BSP site, from reports by BusinessWorld and the Philippine Daily Inquirer on 24 June 2025; the draft itself is dated June 2025.
Does the BSP model risk management draft cover AI?
Yes. Its definition of a model expressly includes methods that apply artificial intelligence and machine learning, and it covers vendor models and models embedded in third-party systems. The BSP's June 2026 AI governance memorandum says the model risk framework will address algorithmic fairness and model risk across the model life cycle.
When will the BSP model risk management rules take effect?
They are not final. The June 2025 draft proposes that institutions comply within two years of the circular's effectivity, with an implementation program in the first year. As of the June 2026 AI memorandum, the BSP still referred to the framework as forthcoming.
| Date | Document | Status |
|---|---|---|
| Jun 24, 2026 | BSP Memorandum No. M-2026-031 — Memorandum No. M-2026-031: Guidance Paper on Governance Principles for Artificial Intelligence (AI) in Financial Services | Final |
| May 30, 2025 | BSP Circular No. 1213 — Circular No. 1213, Series of 2025: Amendments to Regulations on Information Technology Risk Management to Implement Section 6 of the Anti-Financial Account Scamming Act (AFASA) | Final |
| Nov 19, 2026 | BNM AI Feedback Statement — Feedback Statement: Artificial Intelligence (AI) in the Malaysian Financial Sector | Final |
| Nov 19, 2026 | EBA Risk Assessment Report (Nov 2026) — Risk Assessment Report of the European Banking Authority, November 2026, with Box 1 on artificial intelligence related risks and opportunities | Final |
| Nov 4, 2026 | OCC Bulletin 2026-54 — Super Intelligence: Supervisory Framework | Final |
| Nov 4, 2026 | ISA Draft Directive on Technological Means and AI — Draft Directive to Licensees on Providing Services Using Technological Means and Artificial Intelligence (New Version) | Comment period open |
Follow every document these regulators publish
when one of these regulators moves, the next morning's brief says so · six sourced stories · 7 am PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning