You are using an outdated browser and your browsing experience will not be optimal. Please update to the latest version of Microsoft Edge, Google Chrome or Mozilla Firefox. Install Microsoft Edge

June 19, 2025

Thailand Drafts AI Risk Management Guidelines for Financial Service Providers

The Bank of Thailand (BOT) has released draft guidelines establishing principles for managing artificial intelligence (AI) risks in the financial sector. The draft guidelines provide a structured framework for the responsible adoption of AI technologies. Financial service providers will be able to use the guidelines as a reference to appropriately manage their risks in a manner that aligns with internationally recognized best practices.

The BOT is accepting public comments on the draft guidelines until June 30, 2025.

Scope and Application

The draft guidelines apply to all financial service providers, including financial institutions and special financial institutions under the Financial Institution Business Act, as well as payment providers under the Payment Systems Act. These guidelines supplement existing BOT risk management guidelines covering IT risk management, third-party risk management, data governance, and market conduct.

The guidelines define AI systems as systems that mimic human intelligence, including machine learning, deep learning, generative AI (such as large language models), and agentic AI. This definition specifically excludes rule-based automation systems like robotic process automation and condition matching.

Key Risk Management Principles

The guidelines lay out two main principles in managing AI risk.

  1. Governance: Financial service providers should define and establish clear roles and responsibilities for their personnel and AI system supervision structures to uphold FEAT (fairness, ethics, accountability, and transparency) principles as follows:
    • Stakeholder roles and responsibilities. Financial service providers should define roles and responsibilities for boards and executives on AI risk oversight. Responsibilities include establishing an AI system usage policy, designating personnel responsible for AI risk management, and building awareness of AI-related risk within the organization.
    • AI system usage policy. The AI system usage policy should be aligned with organizational objectives, regulatory requirements, and FEAT principles. These policies should be reviewed regularly to respond to technological advancements and evolving risk profiles.
    • Risk management throughout the AI lifecycle. Risk management should encompass the entire AI lifecycle, from establishing risk appetite to implementing continuous risk assessment and control measures tailored to specific use cases. When AI systems are used in strategic functions or customer interactions (e.g., loan approval, account opening), human oversight must be integrated into decision-making processes. In customer interactions with AI systems, customers should be notified and have options to disable or bypass AI features.
  1. Development and security controls: Financial service providers should have risk controls covering the AI development and deployment lifecycle as follows:
    • Data risk. Financial service providers should have measures to assess and ensure the quality, accuracy, currency, volume, and diversity of data used in AI model training. They should also implement data leakage prevention measures.
    • Model development risk. Financial service providers should have (1) clear evaluation metrics for assessing model accuracy and reliability through ongoing testing and monitoring both before and after deployment and (2) measures to ensure the explainability of AI outcomes. For generative AI applications, there should be specific measures to reduce AI hallucination risks.
    • Cybersecurity risk. Financial service providers should have measures to prevent and detect emerging cyber threats targeting AI systems, based on established standards such as the OWASP Machine Learning Security Top 10.

For more details on any aspect of fintech, technology, and cybersecurity in Thailand, please contact Athistha Chitranukroh at [email protected], Nopparat Lalitkomon at [email protected], Pornpan Wichawut at [email protected], Napassorn Lertussavavivat at [email protected], or Rujaporn Paritsantik at [email protected].

RELATED INSIGHTS​ 

June 19, 2024
Vietnam’s financial landscape is set to further transform on July 1, 2024, when the government’s long-awaited Decree No. 52/2024/ND-CP dated May 15, 2024 (“Decree 52”), will officially replace Decree No. 101/2012/ND-CP dated November 22, 2012, on non-cash payments (“Decree 101”). Decree 52 marks an important milestone by introducing the country’s first-ever legal definition of e-money. In addition, the decree brings forth new updates to regulations governing payment and intermediary payment services, laying the groundwork for more comprehensive guidance that will be provided in draft circulars now being developed by the State Bank of Vietnam (SBV). Non-Cash Payment Instruments The new definition of non-cash payment instruments under Decree 52 expands upon the previous definition in Decree 101. Notably, it clearly specifies the issuing entities as payment service providers, financial companies licensed to issue credit cards, and e-wallet service providers. Additionally, the new definition further clarifies that bank cards include debit, credit, and prepaid cards, and adds e-wallets to the list of non-cash payment instruments. Unlawful non-cash payment instruments are still defined as those that are not otherwise specified. E-Money Prior to Decree 52, the concept of e-money lacked a precise legal definition, despite its growing prevalence in forms like prepaid cards and e-wallets. The absence of a clear framework for e-money led to confusion with terms like “cryptpcurrency” and “virtual currency” and left significant ambiguity on whether e-money includes certain instruments, such as online game cards and mobile money. Decree 52 addresses this issue by clearly defining e-money as value in Vietnamese dong (VND) stored electronically and prepaid by customers to banks, foreign bank branches, and e-wallet service providers. It also specifically designates e-wallets and prepaid cards as types of storage mechanisms for e-money. Non-Cash Payment Services Decree 52 categorizes non-cash payment services into services with and without client payment
June 19, 2024
On June 14, 2024, the Personal Data Protection Committee (PDPC) released a draft notification under the Personal Data Protection Act 2019 (PDPA), setting out criteria for how data controllers must delete, destroy, and de-identify personal data. According to the PDPA, a data subject can request that a data controller delete, destroy, or de-identify their personal data in any of the following circumstances: The personal data is no longer necessary for the purposes for which it was collected, used, or disclosed. The data subject has withdrawn their consent for the processing of the personal data, and no other lawful basis for processing remains. The data subject has objected to the processing of their personal data on grounds of legitimate interests or official tasks, the data controller has no other compelling grounds to refuse the request, and the data is not needed for legal claims. The data subject objects to the processing of their personal data for direct marketing purposes. The processing of personal data is unlawful. The draft stipulates that data controllers respond to a data subject’s request to delete, destroy, or de-identify personal data immediately, and within 60 days of receiving the request. If the data controller cannot fulfill the request immediately, they must take interim measures to ensure that the personal data is made difficult to collect, use, or disclose. This includes implementing measures such as preventing access to the data and applying appropriate security measures to protect the data from unauthorized use or disclosure. De-identification or Anonymization of Personal Data In certain circumstances, a data controller may opt to de-identify or anonymize personal data, rather than delete or destroy it. If doing so, the data controller must satisfy the following criteria: There must be a structured process to remove or eliminate all direct identifiers linked to the
June 18, 2024
On June 1, 2024, Thailand’s Securities and Exchange Commission (SEC) issued four notifications amending existing regulations to recognize sustainability-related tokens and institute specific measures for regulating them. These tokens are intended to offer diverse sustainability-related products to ESG funds in Thailand and drive the growth of a sustainable digital economy in the country. The key points in the notifications are summarized below. Definitions Under the notifications, sustainability-related tokens are classified into four types: Green tokens: Digital tokens specifically intended to incentivize or fund projects that promote environmental sustainability. Social tokens: Digital tokens specifically intended to support and fund initiatives that contribute to social welfare. Sustainability tokens: Digital tokens intended to support projects that enhance both environmental and social welfare through funding and incentives. Sustainability-linked tokens: Digital tokens intended to fund activities that promote sustainability. This includes tokens that have adjustable returns based on the performance of the issuing entity or its affiliates in meeting specified sustainability-related goals or outcomes. The offering of sustainability-related tokens is subject to Thailand’s general requirements for token offerings: (1) approval from the SEC and (2) filing the registration statements and the draft prospectus with the SEC before marketing and offering the sustainability-related tokens to public investors in Thailand, unless exempted. The sustainability-related tokens must be offered through an SEC-approved ICO portal, which will assume a role similar to that of a financial adviser and an underwriter in a public offering of securities. Sustainability-Related Token Offerings In addition to complying with the general requirements for token offerings, sustainability-related token offerings must comply with the following measures: Issuer disclosure: The issuer must disclose certain sustainability information, both before and after the offering, according to standards comparable to those of nationally or internationally recognized green, social, and sustainable bonds (GSSBs) and sustainability-linked bonds (SLBs)—such as the principles
June 13, 2024
The Bank of Thailand (BOT) has announced its new Enhanced Regulatory Sandbox, which provides an opportunity to experiment with currently restricted financial innovations under a controlled environment. The BOT is employing a thematic approach to determine the scope of technology or innovations that may participate in the Enhanced Regulatory Sandbox and will only accept applications in each theme for a limited period. The first announced theme is “programmable payments,” which refers to payment and payment-related transactions with automatic execution upon the fulfillment of a predefined condition utilizing distributed ledger technology (DLT) and a smart contract or comparable technology in which electronic data units are issued on an electronic system or network. The application period for programmable payment testing in the Enhanced Regulatory Sandbox runs from June 13 to September 13, 2024. A summary of the programmable payment testing framework under the Enhanced Regulatory Sandbox is provided below. Scope The Enhanced Regulatory Sandbox accepts applications for the following programmable payment activities: Automated payment and settlement upon fulfillment of predefined conditions. Escrow services with predefined delivery or transactional conditions. Asset tokenization through issuance of digital tokens representing rights in an asset, with payment for tokens or payment of benefits or returns to holders of digital tokens occurring automatically when conditions are met. Other testing related to the items mentioned above. Requirements and Conditions Programmable payment testing activities in the Enhanced Regulatory Sandbox must comply with the following requirements and conditions: Electronic data units issued for programmable payment testing must be pegged to the Thai baht (THB) on a one-for-one basis (i.e., 1 unit = THB 1), with the float account storing THB equal to the value of the electronic data units issued. Participants must define the redemption rights of the unitholders and proceed with the THB redemption according to the participants’