Financial institutions remain accountable for third-party AI under new MAS guidelines

Among other expectations, financial institutions will be accountable for AI used in delivered services, including AI developed, operated or provided by third parties.

The logo of the Monetary Authority of Singapore on Feb 13, 2026. (File photo: CNA/Ili Mansor)

This audio is generated by an AI tool.

SINGAPORE: Financial institutions will be expected to assess and manage risks from their use of artificial intelligence, including AI by third-party providers, under new guidelines issued by the Monetary Authority of Singapore (MAS) on Wednesday (Oct 7).

They also remain accountable for AI used in the services they provide even when the technology is developed, operated or supplied by a third party, MAS said.

“Financial institutions should obtain sufficient assurance from third-party providers, assess whether third-party AI is suitable for their intended use, and apply compensating controls where practical constraints or assurance gaps arise,” the central bank said in a media release.

If the risks still cannot be brought within the firm's risk appetite, it should consider limiting, suspending or replacing the third-party AI service.

CNA Games Guess Word Crack the word, one row at a time Buzzword Create words using the given letters Mini Sudoku Tiny puzzle, mighty brain teaser Mini Crossword Small grid, big challenge Word Search Spot as many words as you can Show More Show Less The guidelines will take effect on Oct 7, 2027, with financial institutions allowed to implement them in phases and meet the full requirements by Oct 7, 2028. Under the guidelines, financial institutions are expected to manage AI risks at both the enterprise level and for individual use cases, while building up their capabilities as their use of the technology expands. They should identify their AI use, maintain inventories, assess the risk posed by individual use cases and put in place proportionate controls throughout the AI life cycle. These include data governance, testing, human oversight, cybersecurity, monitoring and change management. MAS said such controls should be reviewed regularly as AI use expands and the technology evolves, pointing to the growing use of agentic AI systems that can operate autonomously and access tools. The authority plans to consult the financial sector in 2027 on what additional guidance on agentic AI would be useful.

The guidelines will take effect on Oct 7, 2027, with financial institutions allowed to implement them in phases and meet the full requirements by Oct 7, 2028.

Under the guidelines, financial institutions are expected to manage AI risks at both the enterprise level and for individual use cases, while building up their capabilities as their use of the technology expands.

They should identify their AI use, maintain inventories, assess the risk posed by individual use cases and put in place proportionate controls throughout the AI life cycle.

These include data governance, testing, human oversight, cybersecurity, monitoring and change management.

MAS said such controls should be reviewed regularly as AI use expands and the technology evolves, pointing to the growing use of agentic AI systems that can operate autonomously and access tools.

The authority plans to consult the financial sector in 2027 on what additional guidance on agentic AI would be useful.

The new guidelines come amid growing scrutiny of the risks posed by increasingly capable and autonomous AI models .

Singapore is studying whether new safety measures are needed in high-risk uses of AI.

Minister for Digital Development and Information Josephine Teo said last week that safeguards put in place by frontier AI companies are insufficient to address the risks posed by increasingly capable systems.

MAS noted on Wednesday that AI technologies are developing rapidly and being adopted in financial services with increasing scale and sophistication.

This includes AI models and systems with greater autonomy in generating outputs, decision-making, or executing actions, the central bank said.

The new guidelines also require boards and senior management to provide effective oversight of AI risks, including by setting out clear roles and responsibilities, risk appetite and risk management frameworks.

Financial institutions will not, however, have to establish a dedicated AI committee if their existing governance structures provide adequate oversight and cross-functional coordination.

The guidelines were established following a public consultation in November 2025.

MAS deputy managing director Ho Hern Shin acknowledged that AI has “significant potential” to improve financial services, but noted that financial institutions need to understand and manage the risks that come with increasingly capable AI systems to reap these benefits sustainably.

“With greater regulatory clarity on financial institutions’ AI usage, financial institutions can innovate with confidence, while maintaining the trust of customers and the resilience of Singapore's financial system,” said Ms Ho.

“MAS will continue to work with the industry to advance sound AI risk management practices in a practical and industry-grounded manner.”

Our chief editor shares analysis and picks of the week's biggest news every Saturday.

Get our pick of top stories and thought-provoking articles in your inbox

Stay updated with notifications for breaking news and our best stories

Join our channel for the top reads for the day on your preferred chat app

除其他要求外，金融机构将对所提供服务中使用的 AI 负责，包括第三方开发、运营或提供的 AI。

新加坡金融管理局的标志，摄于2026年2月13日。（资料照片：CNA/Ili Mansor）

这段音频由人工智能工具生成。

新加坡：新加坡金融管理局（MAS）周三（10月7日）发布的新指南要求金融机构评估和管理使用人工智能（包括第三方供应商提供的人工智能）所带来的风险。

新加坡金融管理局表示，即使人工智能技术是由第三方开发、运营或提供的，他们仍然要对其提供的服务中使用的人工智能负责。

“金融机构应从第三方供应商处获得充分保证，评估第三方人工智能是否适合其预期用途，并在出现实际限制或保证缺口时采取补偿性控制措施，”央行在一份新闻稿中表示。

如果风险仍然无法控制在公司的风险承受范围内，则应考虑限制、暂停或更换第三方人工智能服务。

CNA游戏猜词游戏：逐行破解单词；流行词游戏：用给定的字母组成单词；迷你数独：小巧的谜题，强大的脑力挑战；迷你填字游戏：小网格，大挑战；单词搜索：尽可能多地找出单词。显示更多 显示更少 这些指南将于2027年10月7日生效，金融机构可以分阶段实施，并在2028年10月7日之前满足所有要求。根据指南，金融机构需要在企业层面和具体用例层面管理人工智能风险，并随着技术应用的扩展不断提升自身能力。他们应该明确自身的人工智能应用，维护清单，评估具体用例带来的风险，并在整个人工智能生命周期中实施相应的控制措施。这些措施包括数据治理、测试、人工监督、网络安全、监控和变更管理。新加坡金融管理局（MAS）表示，随着人工智能应用的扩展和技术的演进，应定期审查这些控制措施，并指出能够自主运行和访问工具的智能体人工智能系统正在日益普及。该机构计划在 2027 年就智能体人工智能方面的哪些额外指导意见需要征询金融界的意见。

该指导方针将于 2027 年 10 月 7 日生效，金融机构可以分阶段实施，并于 2028 年 10 月 7 日前满足所有要求。

根据相关指导方针，金融机构应在企业层面和个别用例层面管理人工智能风险，并随着技术应用的扩展不断提升自身能力。

他们应该明确人工智能的使用情况，维护清单，评估各个用例带来的风险，并在整个人工智能生命周期中实施相应的控制措施。

这些包括数据治理、测试、人工监督、网络安全、监控和变更管理。

新加坡金融管理局表示，随着人工智能应用范围扩大和技术发展，此类管控措施应定期审查。他们指出，能够自主运行并访问工具的智能人工智能系统正日益普及。

该机构计划在 2027 年就智能体人工智能方面的哪些额外指导意见需要征询金融界的意见。

新准则的出台，正值人们对日益强大且自主的人工智能模型所带来的风险进行越来越严格的审查之际。

新加坡正在研究在人工智能的高风险应用中是否需要新的安全措施。

新加坡数码发展及新闻部长杨莉明上周表示，前沿人工智能公司采取的保障措施不足以应对日益强大的系统带来的风险。

新加坡金融管理局周三指出，人工智能技术正在快速发展，并在金融服务领域得到越来越大规模和越来越复杂的应用。

该央行表示，这包括在生成输出、决策或执行行动方面具有更大自主性的AI模型和系统。

新准则还要求董事会和高级管理层对人工智能风险进行有效监督，包括明确角色和职责、风险偏好和风险管理框架。

但是，如果金融机构现有的治理结构能够提供足够的监督和跨职能协调，则它们无需设立专门的人工智能委员会。

这些指导方针是在 2025 年 11 月进行公众咨询后制定的。

新加坡金融管理局副局长何赫新承认人工智能在改善金融服务方面具有“巨大潜力”，但他指出，金融机构需要了解和管理日益强大的人工智能系统带来的风险，才能可持续地获得这些好处。

何女士表示：“随着金融机构人工智能使用监管的日益明朗，金融机构可以更有信心地进行创新，同时保持客户的信任和新加坡金融体系的韧性。”

“MAS将继续与业界合作，以务实和符合行业实际情况的方式推进健全的人工智能风险管理实践。”

我们的主编每周六都会分享对本周重大新闻的分析和精选。

订阅我们的邮件，即可获取精选热点新闻和引人深思的文章。

订阅通知，第一时间获取突发新闻和精彩报道。

加入我们的频道，即可在您常用的聊天应用上获取当日热门文章。