UOB's responsible AI
FEAT principles, a tiered governance chain and MAS guidelines.

Last updated Nov 17, 2026 · 10 sources cited

How does UOB govern its use of AI?

UOB governs AI through the Monetary Authority of Singapore's FEAT principles (Fairness, Ethics, Accountability and Transparency), which head of enterprise AI Alvin Eng says UOB helped co-develop as one of the first banks in the consortium in 2018. Every model must go through testing, attestation, evaluation and technical validation before it goes live, he told The Asian Banker at the Singapore FinTech Festival 2025. Oversight is tiered: the AI Ethics and Model Governance Council, then the Data and AI Committee, then the Operational Risk Management Committee (ORMC). The bank monitors fairness metrics such as demographic parity and equal opportunity, uses Shapley values and LIME for explainability, and automates drift monitoring. Its Enterprise GenAI Platform enforces security and compliance guardrails. Externally, UOB is one of 24 participants in MAS's MindForge toolkit (March 2026) and a member of the MAS-ABS taskforce on AI-driven cyber risk (July 2026); MAS's final AI Risk Management Guidelines, issued on October 7, 2026, apply from October 7, 2027. UOB does not publish a stand-alone AI principles document in the sources reviewed.[1][2][3][4][5]

Data as of Nov 17, 2026 · UOB profile updated Nov 17, 2026 · 10 sources cited on this page · 141 banks profiled on this site · Markdown

What bodies and controls govern AI at UOB?

Checked 2026-11-17 · 10 sources
ElementWhat it doesSource date
AI Ethics and Model Governance CouncilInternal body of domain experts at the base of the governance chain, escalating to the Data and AI Committee and then the Operational Risk Management Committee (ORMC)December 2025[1]
AI Ethics and Model Governance TaskforceEmbeds MAS's FEAT principles into every model, as described by Lawrence GohNovember 2025[2]
Pre-launch testing"All our models before they go live, they go through a very rigorous set of testing, attestation, evaluation, and also the model technical validation" (Alvin Eng)December 2025[1]
Fairness and explainabilityMonitors demographic parity and equal opportunity with automated alerts; uses Shapley values, LIME and layer-wise relevance propagation; keeps model cards and data lineageDecember 2025[1]
MLOps and drift monitoringAutomated continuous integration and deployment and automated monitoring for model driftDecember 2025[1]
Enterprise GenAI PlatformEnforces security and compliance guardrails for enterprise-grade generative AI deploymentsNovember 2025[2]
Gen AI reviewReference grounding, human-in-the-loop review and feedback loops on generative AI outputsOctober 2026[6]

How does UOB work with MAS and the industry on AI governance?

InitiativeUOB's roleDate
FEAT principlesOne of the first banks in the consortium to partner with MAS to co-develop the principles2018[1]
Project MindForgeOne of 24 participants in the AI risk-management toolkit covering traditional, generative and agentic AI; provides leadership in the risk and compliance streamsMarch 20, 2026[3][1]
AI-Driven Cyber and Technology Risk Taskforce (ACT)Member of the MAS and ABS taskforce that shares AI cybersecurity use cases and runs proof-of-concept trials of AI-enabled defensive toolsJuly 29, 2026[4]
MAS AI Risk Management GuidelinesApply to UOB from October 7, 2027 for sections 3 and 4 and October 7, 2028 for sections 5 and 6October 7, 2026[5]
Indonesia data rulesCustomer-facing AI applications handling personal data must be hosted onshore; UOB adapts deployment of a standardised central platform to local regulationsNovember 10, 2026[7]

What has UOB not published about AI governance?

The governance chain is described in interviews and partner content rather than in a published policy. The sources reviewed show no stand-alone UOB AI principles document, no count of models under review and no public report of fairness test results. The governing body is called a council in The Asian Banker's interview and a taskforce in The Business Times.[1][2]

Alvin Eng said generative AI models are harder to explain and must therefore be used carefully, and that UOB's AI-specific frameworks sit within its broader model risk management framework.[1]

What does the UOB profile record on this topic?

The rows below are the entries of the UOB AI strategy profile that bear on this page, as of Nov 17, 2026; the profile holds the full record of 17 dated moves, 12 use cases and 21 sources.

Use caseWhat the record showsStatusSource
UOB TMRW personalised insights, nudges and Rewards+ recommenderAI and machine learning drive customised insights and nudges (one in two TMRW customers engage with them) and a merchant recommender model curating more than 1,000 deals for more than one million Rewards+ users; Personetics has powered UOB's personalised insights since 2018 and built the Auto-Save feature.In production[8][9]
Machine-learning name screening and transaction monitoringCo-created machine-learning features in Tookitaki's Anti-Money Laundering Suite, used for name screening and transaction monitoring beyond rules-based systems.In production[10]
ACT taskforce on AI-driven cyber threatsMember of the MAS-ABS taskforce that shares AI cybersecurity use cases, runs proof-of-concept trials of AI-enabled defensive tools and drafts industry guidance.In production[4]
FEAT-aligned Gen AI governance and MindForge participationUOB says its Gen AI enablement follows MAS FEAT principles and Project MindForge, with reference grounding, human-in-the-loop review and feedback loops; it is one of 24 MindForge toolkit participants.In production[6][3]
DateMoveDetailSource
Oct 7, 2026MAS issues final AI Risk Management GuidelinesThe guidelines apply to all financial institutions, including UOB, and take effect on 7 October 2027 for sections 3 and 4 and 7 October 2028 for sections 5 and 6; they cover board oversight, lifecycle controls and accountability for third-party AI.[5]
Jul 29, 2026UOB joins the MAS-ABS taskforce on AI-driven cyber riskMAS and the Association of Banks in Singapore launched the AI-Driven Cyber and Technology Risk Taskforce (ACT), whose members include DBS, OCBC and UOB, to share AI cybersecurity use cases, run proof-of-concept trials of AI-enabled defensive tools and draft guidance.[4]
Mar 20, 2026UOB among 24 participants in MAS MindForge AI risk toolkitThe toolkit covers traditional, generative and agentic AI; participants include DBS, Julius Baer, OCBC, Standard Chartered and UOB, and MAS is setting up an AI risk-management workgroup under BuildFin.ai.[3]
Nov 2, 2022AI Auto-Save on TMRW with PersoneticsUOB launched an AI-driven automated savings feature on TMRW in Indonesia, built with Personetics; the bank said the partnership had delivered more than 150 million personalised insights to customers in four markets since 2018.[9]
Aug 24, 2018Machine learning added to anti-money-laundering checks with TookitakiUOB co-created machine-learning features for Tookitaki's Anti-Money Laundering Suite and was using the system for name screening and transaction monitoring after a six-month pilot, said then head of group compliance Victor Ngo.[10]

Does UOB follow the MAS FEAT principles?

Yes. UOB says its models are governed under MAS's Fairness, Ethics, Accountability and Transparency (FEAT) principles, and head of enterprise AI Alvin Eng says UOB was one of the first banks in the consortium to partner with MAS to co-develop them in 2018.

Who oversees AI risk at UOB?

According to Alvin Eng, oversight starts with the AI Ethics and Model Governance Council, an internal body of domain experts, escalates to the Data and AI Committee and ultimately reaches the Operational Risk Management Committee (ORMC). Model owners always come from the business.

How does UOB check AI models for bias?

UOB monitors fairness metrics such as demographic parity and equal opportunity with automated alerts and dashboards, and uses Shapley values, LIME and layer-wise relevance propagation for explainability, Alvin Eng told The Asian Banker.

When do the MAS AI risk guidelines apply to UOB?

MAS issued its final AI Risk Management Guidelines on October 7, 2026. Sections 3 and 4 take effect on October 7, 2027 and sections 5 and 6 on October 7, 2028, for all financial institutions including UOB.

  1. UOB strengthens responsible AI governance for enterprise deployment
    The Asian Banker · Dec 3, 2025
  2. Engineering the future of banking across Asean (Future of Finance 2025)
    The Business Times · Nov 12, 2025
  3. MAS launches AI risk toolkit for financial institutions with case studies from DBS, peers
    The Business Times · Mar 20, 2026
  4. MAS, ABS launch finance sector taskforce to tackle AI-driven cyber risks
    Retail Banker International via Yahoo Finance · Jul 29, 2026
  5. MAS rolls out AI risk management guidelines for Singapore's financial sector
    The Business Times · Oct 7, 2026
  6. Gen AI Enablement | UOB @ SFF 2025 (undated page, retrieved 2026-10-24)
    United Overseas Bank · Oct 24, 2026
  7. What this S'pore bank weighs in tailoring its AI tools across Asean
    The Business Times · Nov 10, 2026
  8. UOB Annual Report 2025
    United Overseas Bank · Mar 19, 2026
  9. UOB and Personetics launch AI-driven automated savings feature on its TMRW mobile banking app – the first of its kind in Southeast Asia
    United Overseas Bank · Nov 2, 2022
  10. UOB, Tookitaki team up on machine learning to fight money laundering
    The Business Times · Aug 24, 2018

When UOB moves on AI, you'll read it here first.

when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am PT · free

plus every regulation, bank and agent page update, the morning after · leave any morning