On 9 July 2026, the Electronic Transactions Development Agency (ETDA) released the Draft Artificial Intelligence Act (the “Draft AI Act”) for public consultation. If enacted, it would establish Thailand’s first comprehensive AI regulatory framework. The Draft AI Act adopts a risk-based approach covering AI developers, providers, deployers, and platform operators, including certain overseas entities whose AI activities affect individuals in Thailand. AI systems would be classified according to risk, with different requirements applying to prohibited, high-risk, licensed, and transparency-regulated AI systems. The consultation is scheduled to close on 14 August 2026.
First, businesses will need to assess the classification and risk level of their AI systems and implement appropriate compliance measures. The Draft AI Act also introduces requirements concerning training data and personal data. Regulators may, in certain circumstances, obtain technical information relating to AI development, including training, validation, and testing data. Businesses participating in AI sandbox programmes may also receive access to government data subject to safeguards for personal data and data subjects’ rights.
Second, new transparency obligations would apply to AI-generated content. Developers and platform operators may be required to identify or label AI-generated content, including deepfakes and chatbot-generated material. Users and employees may also be required to disclose AI use when publishing AI-generated or modified content concerning specified sensitive matters. Businesses should therefore consider implementing internal policies governing the use and disclosure of AI-generated content.
Third, overseas AI providers and vendors serving government agencies or critical information infrastructure (CII) may face additional requirements. Certain providers may be required to appoint a local representative, while AI services may be subject to mandatory contractual terms and data localisation requirements. Businesses should therefore review their existing technology, cloud, data-processing, and cross-border transfer arrangements.
Finally, the Draft AI Act would provide regulators with broad powers to investigate serious AI incidents, suspend services, and order recalls. It also proposes a strict and potentially joint liability regime, meaning responsibility for AI-related harm may extend across developers, providers, deployers, and platform operators. From a broader legal perspective, businesses should begin mapping their AI systems, reviewing contractual risk allocation and data governance arrangements, and preparing appropriate AI governance frameworks. Although the legislation remains subject to consultation and may be revised, early preparation will help businesses manage compliance risks as Thailand moves towards comprehensive AI regulation.
Thailand’s Draft AI Act Introduces New Compliance Requirements for Businesses_Bangkok Global Law