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

Subscribe to Updates

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 Professionals

Industries

Fintech

Technology

Practices

Banking and Finance

Location

RELATED INSIGHTS​ 

December 7, 2023
Thailand’s Securities and Exchange Commission (SEC) has updated its regulations regarding public initial coin offerings (ICOs) that exhibit debt-like or infrastructure-backed characteristics. These new regulations are set to transform the landscape for digital fundraising, ensuring enhanced investor protection and risk mitigation. The introduction of these regulations is a major move toward cultivating a more transparent and secure digital token market in Thailand. It highlights the importance of comprehensive information disclosure and aligning digital-asset fundraising practices with established financial standards. The adjustments require more collaboration between token issuers and ICO portals to comply with the developing regulatory environment. The key features of the new regulations, which took effect on November 16, 2023, are summarized below. Debt-Like ICOs Debt-like digital tokens are digital tokens for which a fixed rate of return is predetermined. The rate does not vary based on the performance of the project or enterprise, but it may be supplemented by extra returns. To ensure that investors have access to adequate and essential information prior to making informed investment decisions, and to evaluate the risk of default by the issuer, the regulation requires disclosure of the project’s creditworthiness and risk assessment factors, as detailed below. Project creditworthiness. Disclosures about project creditworthiness should include information about the credibility assessment of the project, which should be conducted reasonably by an ICO portal or an independent expert. Risk assessment factors. Disclosures should include information about the factors considered in the assessment of risks—particularly the financial capability to repay the investment project’s debt. Infrastructure-Backed ICOs Infrastructure-backed digital tokens are issued for investment in infrastructure assets or to earn a share of the revenue stream from infrastructure projects. The regulations for the ICOs of these infrastructure-backed tokens have been revised to align with those for infrastructure trusts. These regulatory updates cover various aspects: Definition
Read More
November 27, 2023
The emergence of generative artificial intelligence (AI) has transformed the landscape for innovators and creators. As many legal practitioners have pointed out, it’s imperative for both developers of AI and artists using generative AI to understand the intricacies of intellectual property (IP) strategies so they can navigate this evolving terrain successfully. This article lays out some essential considerations relating to the major types of IP for both developers and creators in the realm of generative AI. IP Strategies for Developers of Generative AI Developers of generative AI technologies play a pivotal role in the innovation landscape. There are three overarching IP-related issues to consider: protecting their own intellectual property, mitigating the risk of violating other people’s IP rights, and IP commercialization. Key aspects of these concerns, along with suggested approaches for developers, are outlined below. Protecting IP Copyrights. One of the primary considerations for AI developers is the protection of AI-generated works, such as art and source code. The good news is that in most countries, these creations enjoy copyright protection without the need for registration. As a result, the works are automatically protected from the moment of creation. However, it’s crucial to maintain comprehensive records of your work to establish your ownership. Trademarks. Trademarks are vital for AI developers looking to establish and protect their brand. Pay close attention to Nice classifications, particularly class 9 (for software), class 35 (for business management and online marketing), and class 42 (for software design and development). Registering trademarks in these classes can provide robust protection for your brand and products. Patents. For truly innovative AI algorithms, techniques, or processes, consider the option of patenting. Patents offer strong protection, but they require a thorough application process and the documentation of your innovation, including evidence that the invention is novel, non-obvious, and practically
Read More
November 27, 2023
Thailand’s Electronic Transaction Development Agency (ETDA) has released two new subordinate regulations under the Royal Decree on Digital Platform Services: one detailing the assessment of digital platform services (DPSs) that will be deemed “high-risk” and subject to additional obligations, and another setting guidelines on user verification and authentication for all DPSs. The two subordinate regulations are summarized below. Impact Assessment of DPS Operations Under the Royal Decree on Digital Platform Services, DPS operations that have the risk of seriously impacting financial and commercial security, reliability and credibility of data message systems, or the general public are subject to additional obligations. The first subordinate regulation mentioned above (officially titled Notification of the Electronic Transactions Commission Re: Criteria for Impact Assessment on Operation of Digital Platform Services) outlines the criteria for the ETDA to determine which DPSs are “high-risk.” DPSs falling under this designation include: DPSs whose total value of transactions conducted through the platform in Thailand exceeds THB 100 million (approx. USD 2.8 million) per year; DPSs whose operators have not registered their entities with the Department of Business Development (DBD)—notably overseas operators—and that have 100 or more merchants or business users in Thailand or total users in Thailand between 5 and 10 percent of the country’s population (i.e., approx. 3.3–6.1 million users, calculated using official 2022 figures); DPSs that allow their users to freely post certain messages, or do certain acts, that may affect the public in certain cases, such as: (1) unlawful messages or acts; (2) messages or acts that may affect a child’s rights or people’s fundamental rights; and (3) messages or acts that may negatively affect political opinions of Thai citizens (whether before or after an election) or statements or actions likely to negatively affect other individuals due to gender differences or sexual violence. After considering
Read More
November 23, 2023
On November 14, 2023, Thailand’s Personal Data Protection Committee (PDPC) published a draft notification on collection of personal data regarding criminal records. The draft notification aims to provide clarifications and prescribe further criteria for processing criminal record data under the Personal Data Protection Act (PDPA), which generally requires the processing of criminal records to be carried out under the control of the relevant official authority under the law or under a data protection measure implemented according to rules prescribed by the PDPC. After its eventual passage, the draft notification will have important implications for businesses’ recruitment and human resources activities in relation to individuals with criminal records. Key aspects of the draft notification include the following: “Personal data regarding a criminal record” and “criminal record data” denote personal data related to the investigations of criminal offenses, criminal prosecution, or criminal punishment that is official information or certified by the relevant supervisory authority, regardless of whether that action is connected to a final judgment. Under the draft notification, data controllers may process criminal record data for the purpose of a recruitment process, checking the qualifications of personnel, and considering the suitability of a person for a position if the processing activities are required by law or when a data controller obtains explicit consent from the data subject. Furthermore, the necessity of processing the criminal record data must be announced at the beginning of the recruitment process. Data controllers’ requests for explicit consent to collect a data subject’s criminal record data must also notify the data subject of the consequences of not providing consent or withdrawing consent. The draft notification sets the allowable retention period for criminal record data at a maximum of six months from the end of the processing activities specified above. After the retention period ends, the criminal
Read More