TheMonetary Authority of Singapore (MAS) has drawn a clearer regulatory linearound the use of artificial intelligence in financial services. On 7 October2026, the regulator published its Guidelines on Artificial Intelligence RiskManagement, establishing supervisory expectations for financialinstitutions using AI, from traditional machine-learning models to GenerativeAI and increasingly autonomous AI agents. The guidelines follow a publicconsultation launched in November 2025 and will begin taking effect on 7October 2027.AI isincreasingly moving beyond chatbots and productivity tools into areas that caninfluence financial decisions, customer interactions, compliance, riskmanagement and potentially execution itself. MAS specifically warns that AIagents with access to tools could autonomously perform erroneous orunauthorised actions. It also identifies financial, operational, conduct,financial crime and reputational risks arising from AI, alongside security,privacy and third-party risks associated particularly with Generative AI.AFramework Built Around Risk, Not a Ban on AITheframework is not designed to slow the adoption of AI. Its central idea isproportionality. MAS expects financial institutions to understand where AI isbeing used, determine how important and risky each application is, and thenapply controls that match that risk. A tool used to proofread an internal emailclearly does not require the same controls as an AI model influencinginvestment recommendations or customer risk assessments.London's trading industry is coming home!Thisdistinction is important when considering whether the framework is"mandatory". These are guidelines setting out MAS' supervisoryexpectations, rather than a standalone piece of legislation imposingidentical technical requirements on every company. Yet regulated firms shouldnot interpret this as optional guidance. MAS states that all financialinstitutions should adhere to its high-level expectations when deploying AI,while implementation should be proportionate to their size, activities and AIrisk exposure. The guidelines also complement existing MAS requirements andprinciples, including Fairness, Ethics, Accountability and Transparency, knownas FEAT.This givesfirms roughly one year to establish the core governance layer and another yearto complete the broader operational framework. MAS has also said that itintends to consult the industry in 2027 on whether further guidancespecifically addressing agentic AI would be useful.WhoNeeds to Follow the Guidelines?The scopeis broad. MAS says the guidelines apply to all financial institutions and allforms of AI technology. The categories listed by the regulator include banks,fund managers, financial advisers, insurers, payment institutions, exchanges,clearing houses, custodians, credit rating agencies and firms dealing incapital markets products. Exempt futures and OTC derivatives brokers are alsoexplicitly included among the categories covered.Read fullanalysis on Finance Magnates Intelligence portal:This article was written by Sylwester Majewski at www.financemagnates.com.