# BNM AI Discussion Paper: Discussion Paper on Artificial Intelligence in the Malaysian Financial Sector

Source: https://www.bankingnewsai.com/ai-regulation/documents/bnm-ai-discussion-paper-2025
Last updated: Nov 20, 2026

Bank Negara Malaysia issued the Discussion Paper on Artificial Intelligence in the Malaysian Financial Sector on 5 August 2025 and took written feedback until 17 October 2025. It set out BNM's view that 'the existing regulatory framework remains broadly adequate in addressing risks posed by AI at this stage', because most AI use is in non-financial risk areas and is unlikely to fully replace human decision making for critical tasks. It expects financial service providers to manage AI across the model lifecycle with controls proportionate to complexity, risk and intended use, and it backed the banking industry's own AI Governance Framework. The paper is non-binding; BNM answered it with a feedback statement on 19 November 2026.

## At a glance

| Field | Value |
| --- | --- |
| Authority | [Malaysia (BNM / SC)](https://www.bankingnewsai.com/ai-regulation/malaysia) |
| Type | Consultation |
| Status | Final |
| Published | Aug 5, 2025 |
| Comment deadline | Oct 17, 2025 |
| Applies to | Financial service providers regulated by BNM: licensed banks, Islamic banks and investment banks, insurers and takaful operators, development financial institutions and e-money issuers. A non-binding publication that, in BNM's words, 'may not result in an exposure draft or policy document under the same title' |
| Official text | https://www.bnm.gov.my/documents/20124/938039/DP_Artificial_Intelligence_in_the_Financial_Sector.pdf |

## Key points

- Definition: for the paper's purposes AI is 'the use of advanced computer systems capable of processing and analysing large volumes of data and performing tasks that traditionally require human intelligence, including generating content or making predictions to aid in decision making processes'. BNM asked whether a formal sector-specific regulatory definition would help.
- Adoption (BNM AI Survey 2024): 71% of banking institutions and development financial institutions had implemented at least one AI application in 2024, up from 56% the year before; 77% of insurers and takaful operators, up from 58%. Customer analytics and marketing (28% of reported projects) and internal operational improvements (22%) led; pricing, scoring and underwriting accounted for 5%.
- Risks named: bias and discrimination in credit underwriting from flawed data, model opacity, misuse of model outputs without compensating model risk management controls, dependence on a small set of foundation-model developers, data leakage, GenAI hallucinations in customer chatbots, data poisoning in retrained models, herding from shared models and datasets, and AI-enabled scams including deepfakes and mass e-KYC attacks.
- Regulatory approach: parity, proportionality and neutrality. Existing technology-agnostic, outcome-focused requirements apply to AI, but BNM 'may explore the need to introduce new regulatory expectations' and may escalate supervisory focus if AI-related risks become more material, for example through widespread use in critical functions or to replace human decision making.
- Model risk: a November 2024 survey by the Chief Risk Officers' Forum found that 11 of 14 banks with AI applications had model risk management frameworks that apply to AI models; the paper calls validation a particularly critical step and reports wider use of explainable-AI techniques, champion-challenger models, bias detection and fairness metrics among the most active adopters.
- Data: providers are expected to handle personally identifiable information responsibly, stay compliant with existing and emerging requirements including the Personal Data Protection Act as amended in 2024, and practise transparency in data sourcing, informed consent and sound data management.
- Industry principles: BNM commended the AI Governance Framework drafted by the Chief Risk Officers' Forum of the Asian Institute of Chartered Bankers, built on fairness, ethical use, accountability, transparency, explainability, reliability and security, and encouraged insurers and other segments to align with it or develop complementary guidance.
- Development: BNM asked providers to prioritise 'win-win-win' use cases such as AI fraud detection and AML monitoring, offered its Regulatory Sandbox for AI use cases that face regulatory impediments, and proposed industry-wide work on AI risk management guidelines, model validation knowledge-sharing, talent and consumer awareness.

## What changed for banks

The paper was BNM's first published regulatory and development approach to AI, following a 2021 preliminary survey of AI and machine learning use, a 2023 box article in its Financial Stability Review on early adoption and industry-wide AI surveys from 2023. It chose existing, technology-neutral requirements over AI-specific rules, placed the voluntary National Guidelines on AI Governance and Ethics issued by the Ministry of Science, Technology and Innovation in 2024 as the national backdrop, and left principle-setting to the industry. The consultation closed on 17 October 2025; BNM's feedback statement of 19 November 2026 kept the same approach. The PDF is dated August 2025; BNM's release page (https://www.bnm.gov.my/-/dp-aifs25) gives 5 August 2025 as the date of issuance and 17 October 2025 as the feedback deadline.

## Use cases it governs

- [AI governance (general)](https://www.bankingnewsai.com/ai-regulation/by-use-case#governance-general)
- [Model risk management](https://www.bankingnewsai.com/ai-regulation/by-use-case#model-risk)
- [Generative & agentic AI](https://www.bankingnewsai.com/ai-regulation/by-use-case#generative-agentic-ai)
- [Credit scoring & underwriting](https://www.bankingnewsai.com/ai-regulation/by-use-case#credit-underwriting)
- [Fair lending & discrimination](https://www.bankingnewsai.com/ai-regulation/by-use-case#fair-lending)
- [Third-party & vendor AI](https://www.bankingnewsai.com/ai-regulation/by-use-case#third-party-vendors)
- [Data & privacy](https://www.bankingnewsai.com/ai-regulation/by-use-case#data-privacy)
- [Fraud detection](https://www.bankingnewsai.com/ai-regulation/by-use-case#fraud)
- [Cybersecurity](https://www.bankingnewsai.com/ai-regulation/by-use-case#cybersecurity)

## FAQ

### Is the BNM AI discussion paper binding?

No. It describes itself as a non-binding publication that may not result in an exposure draft or policy document under the same title. BNM's feedback statement of 19 November 2026 confirmed that BNM is not introducing AI-specific regulation at this stage, while existing policy documents such as RMiT and Fair Treatment of Financial Consumers continue to apply to AI use.

### What does BNM expect banks to do about AI model risk?

The paper expects financial service providers to develop holistic risk management and oversight across the lifecycle of AI models, with controls proportionate to the complexity, potential risks and intended use of each model. It cites an industry survey in which 11 of 14 banks with AI applications already applied model risk management frameworks to AI models, and names validation as a particularly critical step.

### When did consultation on the BNM AI discussion paper close?

Written feedback was due by email to BNM by 17 October 2025. BNM received feedback from 113 respondents, about 80% of them financial service providers, and published its Feedback Statement on Artificial Intelligence in the Malaysian Financial Sector on 19 November 2026.

## Related documents

- [BNM AI Feedback Statement](https://www.bankingnewsai.com/ai-regulation/documents/bnm-ai-feedback-statement-2026) — Feedback Statement: Artificial Intelligence (AI) in the Malaysian Financial Sector (Nov 19, 2026)
- [BNM Governor AICB Nexus 2026 speech](https://www.bankingnewsai.com/ai-regulation/documents/bnm-governor-aicb-nexus-speech-2026) — Governor's Opening Address at the Inaugural AICB Nexus 2026: Advancing Banking, Governance and Financial Innovation in Malaysia and Asia (Jul 8, 2026)
- [BNM Risk Management in Technology (RMiT)](https://www.bankingnewsai.com/ai-regulation/documents/bnm-rmit-2025) — Risk Management in Technology (RMiT) policy document (Nov 28, 2025)
- [BNM e-KYC Policy Document](https://www.bankingnewsai.com/ai-regulation/documents/bnm-ekyc-2024) — Electronic Know-Your-Customer (e-KYC) policy document (Apr 15, 2024)
- [BNM Fair Treatment of Financial Consumers](https://www.bankingnewsai.com/ai-regulation/documents/bnm-ftfc-2024) — Fair Treatment of Financial Consumers policy document (Mar 27, 2024)
- [SC Guidelines on Technology Risk Management](https://www.bankingnewsai.com/ai-regulation/documents/sc-gl-2-2023-technology-risk-management) — Guidelines on Technology Risk Management (Securities Commission Malaysia) (Aug 1, 2023)
- [BoE/PRA/FCA effective practices on AI cyber risk (Nov 2026)](https://www.bankingnewsai.com/ai-regulation/documents/uk-effective-practices-cyber-risk-ai-2026) — Effective practices: Cyber risk and resilience management relating to artificial intelligence (AI) (Nov 19, 2026)
- [EBA Risk Assessment Report (Nov 2026)](https://www.bankingnewsai.com/ai-regulation/documents/eba-risk-assessment-report-autumn-2026) — Risk Assessment Report of the European Banking Authority, November 2026, with Box 1 on artificial intelligence related risks and opportunities (Nov 19, 2026)

Last reviewed Nov 20, 2026. Cite the official text (https://www.bnm.gov.my/documents/20124/938039/DP_Artificial_Intelligence_in_the_Financial_Sector.pdf) for the rule and this page for the summary and dates.

---

Canonical page: https://www.bankingnewsai.com/ai-regulation/documents/bnm-ai-discussion-paper-2025
Part of [BankingNewsAI](https://www.bankingnewsai.com/) — a free daily brief on AI in banking, an AI regulation tracker (60 authorities, 360 documents) and AI-strategy profiles of the 120 largest US banks. Markdown versions of every reference page: append `.md` to the page URL; index at https://www.bankingnewsai.com/llms.txt.
