Support Centre

You have out of 5 free articles left for the month

Signup for a trial to access unlimited content.

Start Trial

Continue reading on DataGuidance with:

Free Member

Limited Articles

Create an account to continue accessing select articles, resources, and guidance notes.

Free Trial

Unlimited Access

Start your free trial to access unlimited articles, resources, guidance notes, and workspaces.

Thailand: Update on proposals for AI regulations

The rapid ascent of artificial intelligence (AI) has paved the way for a new era of innovation and is reshaping our daily lives. The emergence of generative AI, a content-generating tool, is a recent example of how quickly these developments can take place. However, they increasingly challenge the applicability of current laws, demanding tailored regulations. Kritiyanee Buranatrevedhya and Burin Saekow, from Baker & McKenzie LLP, discuss the proposed updates to manage these changes.  

MF3d / Signature collection / istockphoto.com

Current regulatory landscape 

According to the recent development of AI, which can sometimes operate without human intervention, and the legal status of AI itself, existing laws may struggle to address emerging issues. For instance, AI-generated content may contain copyrighted data and personal data scraped from the internet, posing challenges for AI developers to determine the legal ground for data mining and the use of personal data as well as for rights holders to enforce their rights. Therefore, the adequacy of current regulations for safeguarding stakeholders in the AI ecosystems remain uncertain as there has been no specific AI laws in Thailand before.  

Recent developments in AI regulation 

In 2022, Thai government entities began studying the regulatory development and approaches of several jurisdictions in regulating AI, and, as a result, two different draft legislations were introduced by two different regulators. 

The Draft Royal Decree on Business Operations that Use Artificial Intelligence System 

The Draft Royal Decree on Business Operations that Use Artificial Intelligence System (the Draft AI Royal Decree) was first introduced by the Office of the National Digital Economy and Society Commission (ONDE) in public hearings held in late 2022. It will be enacted under the Electronic Transaction Act B.E. 2544 (2001) (ETA) which gives the power to issue such Decree. 

Key regulatory principles 

The Draft AI Royal Decree is influenced by the EU AI Act, which proposes a risk-based regulatory approach; that is, the regulatory requirements and oversight for AI systems will be stricter as the risk of such a system increases. 

Overview of specific regulatory frameworks 

AI systems are categorized into three categories, with key obligations varying for each type of regulated AI. For example: 

  • prohibited AI systems, such as AI that exploits any of the vulnerabilities (e.g., disability and children), would be forbidden from being put into service except in limited cases; 
  • high-risk AI systems, such as AI in critical infrastructure, medical devices, and autonomous vehicles, could be required to register prior to the provision of service and compliance with some obligations; and 
  • AI systems with limited risk, such as chatbots, AI used for impersonation, and deepfakes, would be subject to transparency obligations and communications to relevant parties.  

Penalties 

Administrative orders, including administrative fines, suspension of the service, and compensation, could be imposed, as well as criminal liability under ETA, if conducting business without registration.  

The Draft Act on the Promotion and Support of AI Innovations in Thailand

The Draft Act on the Promotion and Support of AI Innovations in Thailand (Draft AI Act) was first introduced by the Electronic Transactions Development Agency (ETDA) in early 2023 and was later updated and placed for public hearing again in mid-2023. Unlike the Draft AI Royal Decree, the Draft AI Act intends to be enacted as a separate Act and is not subject to ETA. 

Key regulatory principles 

The aim of the Draft AI Act is to promote and provide necessary support to expedite the AI ecosystem through proposed mechanisms, which will be further described in the summary below. 

Overview of specific regulatory frameworks 

There are several key proposed mechanisms, such as:  

  • AI business registration: AI operators who register with the ETDA will be eligible to receive certain benefits, such as AI Clinic; 
  • AI sandbox: The AI sandbox will allow AI operators to test their AI systems in a controlled environment and seek the regulator's recommendations; 
  • standard contract terms: There are some required minimum contents to be disclosed to the public for transparency and to prevent unfair discriminatory contract clauses;  
  • risk assessment: Regulators may prescribe criteria and methods for conducting a risk assessment of an AI system based on AI ethics principles as guidance for certain AI stakeholders; and  
  • AI systems that can be operated under strict supervision: Regulators may prescribe the list of such AI systems where it could extend to the offshore AI operator offering service to the user in Thailand, with some additional obligations.  

Penalties 

The ETDA will be empowered to issue a corrective order to AI service providers that do not act according to the necessary quality of service. 

What's next? 

Both draft laws have varying aims for AI regulations: the Draft AI Royal Decree suggests stricter rules, whereas the Draft AI Act leans towards empowering AI development while safeguarding consumers. Yet, in these stages of AI legislation, their legislative plans remain uncertain. It is advisable for AI service providers, operators, and/or developers to closely monitor the evolution of these AI regulations, as it would significantly reshape Thailand's AI regulatory landscape. 

Kritiyanee Buranatrevedhya Partner 
[email protected]  
Burin Saekow Associate 
[email protected] 
Baker & McKenzie LLP, Bangkok