Under the new rule, firms must submit detailed documentation on model design
The Monetary Authority of Singapore (MAS) has announced that all artificial intelligence and machine learning applications used by financial technology firms operating in the city-state must undergo independent review before deployment. The directive applies to both in-house and third-party AI systems, with MAS emphasizing that reliance on external vendors does not absolve financial institutions of responsibility for risks arising from AI failures. The policy takes effect immediately and covers use cases ranging from credit scoring to fraud detection and algorithmic trading. MAS tightens oversight amid growing AI adoption in finance MAS stated that the rapid integration of AI into financial services has introduced new operational, ethical, and systemic risks that existing regulatory frameworks may not fully address. The authority cited past incidents where flawed third-party AI models led to incorrect lending decisions or biased outcomes, noting that financial institutions cannot shift blame to suppliers when things go wrong.
Breaking news
Town AI Secures $108 Million Series B, Valued at $1 Billion After $55 M Pre‑Series
Apple Prepares for Major Product Launch Event on October 28
Apple Releases Release Candidates for watchOS, tvOS and visionOS 27.2
Apple Limits iPhone Duo and iPhone 18 Pro Max to Reservation-Only Sales in Japan and South KoreaUnder the new rule, firms must submit detailed documentation on model design, data sources, testing procedures, and governance controls to an approved independent reviewer. MAS will maintain a list of accredited assessors and may conduct spot checks to ensure compliance. The move aligns with Singapore’s broader strategy to foster innovation while safeguarding financial stability and consumer protection. How will independent reviews change AI deployment in FinTech? The requirement for independent assessment is expected to slow down the rollout of some AI tools, particularly for smaller FinTechs lacking resources for rigorous validation. However, MAS argues that this delay is necessary to prevent costly errors and reputational damage. Industry experts suggest the policy could encourage greater standardization in AI development practices and improve transparency across the sector. Some firms have welcomed the clarity, saying it helps build trust with regulators and customers alike.
MAS emphasized that the goal is not to stifle innovation but to ensure that AI
MAS emphasized that the goal is not to stifle innovation but to ensure that AI applications are safe, fair, and accountable before they reach the market. Frequently Asked Questions What types of AI systems are covered by the new MAS rule? The rule applies to all AI and machine learning use cases in financial services, including credit assessment, fraud detection, customer chatbots, algorithmic trading, and risk modeling, whether developed internally or sourced from third-party vendors. Who will conduct the independent reviews of AI models? MAS will accredit independent reviewers with expertise in AI validation, financial risk, and data ethics. Firms must select from this approved list, and MAS reserves the right to audit the review process itself. What happens if a firm deploys an AI system without independent approval?
MAS warned that deploying unreviewed AI systems could result in supervisory action, including fines, restrictions on business activities, or requirements to remediate and retest the model under regulatory supervision.
