Thailand Draft Artificial Intelligence Act
ETDA draft released for consultation 9 July 2026: prohibited AI, high-risk AI, AI subject to licensing, and transparency duties for deepfakes, chatbots and generative AI, with a national AI regulator.
| Jurisdiction | Thailand |
|---|---|
| Category | AI Regulations |
| Status | Proposed |
| Latest development |
Analysis
Thailand’s Draft Artificial Intelligence Act is still a draft under public consultation, so the requirements below should be treated as proposed obligations, not finalized law. The consultation was opened by Thailand’s Electronic Transactions Development Agency (ETDA) in early July 2026 and reportedly ran until 14 August 2026.Thailand draft AI act commentary, Thailand draft AI act update
Key Requirements
- Risk-based classification of AI systems: the draft reportedly divides AI into prohibited AI, high-risk AI, AI subject to licensing, and AI subject to transparency duties for uses such as deepfakes, chatbots, and generative AI.Thailand draft AI act commentary, Thailand tech governance analysis
- Prohibited AI uses: the draft framework is described as banning certain AI practices, especially where systems create unacceptable risks such as discriminatory profiling or covert manipulation.Thailand tech governance analysis, Thailand draft AI act commentary
- High-risk AI obligations: the draft is described as imposing stronger governance, risk-control, and supervision requirements on high-risk AI systems, with more intensive obligations than for lower-risk systems.Thailand draft AI act update, Thailand regulatory landscape PDF
- Licensing for certain AI activities: the draft reportedly requires licensing for some AI activities or operators, indicating a controlled regime for higher-impact deployments.Thailand draft AI act commentary, Thailand tech governance analysis
- Transparency duties for AI-generated content: obligations reportedly apply to deepfakes, chatbots, and generative AI, including disclosure or labeling duties when content is AI-generated.Thailand draft AI act commentary, Thailand tech governance analysis
- Expanded extraterritorial reach: the draft is described as applying to developers, providers, deployers, and platform operators, including offshore entities serving Thailand.Thailand draft AI act commentary, Thailand AI regulation overview
- National AI governance structure: commentary indicates the draft contemplates a national AI regulator or empowered AI governance body, together with enforcement powers over AI systems.Thailand tech governance analysis, Thailand AI regulation overview
- Enforcement powers: the draft is reported to authorize corrective measures such as service suspension, deployment blocking, and potentially recall measures for noncompliant systems.Thailand draft AI act commentary, Thailand tech governance analysis
Compliance Challenges
- Draft status and moving target risk: organizations face the challenge that the law is not final, and obligations may change after consultation and legislative revision.Thailand draft AI act update, Thai-Ukraine legal framework update
- Risk classification uncertainty: companies will need to determine whether a system is prohibited, high-risk, licensable, or merely subject to transparency duties, which can be difficult for hybrid or multifunction systems.Thailand draft AI act commentary, Thailand tech governance analysis
- Cross-border compliance: offshore providers and global platform operators may need Thai-specific governance, legal representation, or localization-style controls if the draft’s extraterritorial features are retained.Thailand draft AI act commentary, Thailand AI regulation overview
- Content labeling and provenance controls: deepfake and generative-AI disclosure rules require technical mechanisms for labeling, watermarking, or user disclosure, which can be difficult at scale.Thailand tech governance analysis, Thailand draft AI act commentary
- Governance for third-party models: businesses integrating vendor models must manage uncertainty over responsibility allocation, especially where the draft contemplates strict or joint liability.Thailand draft AI act commentary, Thailand AI regulation overview
- Example of industry concern: legal and business commentary on the consultation noted concerns that stricter AI controls could affect investment sentiment and delay deployment decisions for AI products in Thailand.Thailand business reaction report, Thailand draft AI act commentary
Implementation Best Practices
- Build an AI inventory first: document every AI system, use case, vendor, data source, user group, and deployment geography so you can classify systems against the draft’s tiers.Thailand draft AI act commentary, Thailand regulatory landscape PDF
- Adopt a risk management framework: use established frameworks such as the NIST AI Risk Management Framework and ISO/IEC 42001 to operationalize governance, testing, incident response, and accountability.NIST AI RMF, ISO/IEC 42001 overview
- Map duties by risk tier: create a controls matrix that maps prohibited, high-risk, licensable, and transparency-only systems to required approvals, testing, monitoring, and disclosure tasks.Thailand tech governance analysis, Thailand draft AI act update
- Implement content provenance controls: prepare labeling, watermarking, and metadata retention for generative AI output, especially for chatbots and deepfake-related use cases.Thailand tech governance analysis, C2PA specification
- Create vendor governance terms: include audit rights, incident notification, model-change notice, data-use limits, and responsibility allocation in AI procurement contracts.Thailand draft AI act commentary, ISO/IEC 42001 overview
- Use structured assessment tools: apply algorithmic impact assessments and model cards before launch, and maintain documented testing for bias, safety, and explainability where relevant.NIST AI RMF, OECD AI Principles
- Track regulatory consultation materials continuously: monitor ETDA announcements and official consultation documents so controls can be updated before enactment.ETDA homepage, Thailand draft AI act update
Recent Updates
- 9 July 2026: ETDA reportedly opened public consultation on the draft AI Act.Thailand consultation notice summary, Thailand draft AI act commentary
- 2 July 2026: a revised draft was reported as having been released for public hearing, with comments expected over roughly 30 days.Thailand draft AI act update, Thai-Ukraine legal framework update
- 14 August 2026: multiple sources report the consultation closed on this date.Thailand draft AI act commentary, Thailand consultation notice summary
- Phased implementation: commentary states that some core product-launch measures would take effect immediately upon publication, while risk-control and incident-related provisions would take effect 180 days after publication if the draft proceeds in that form.Thailand draft AI act update
- No final effective date yet: because the measure is still a draft, there is no final compliance deadline until the legislative process is completed.Thai-Ukraine legal framework update, Thailand draft AI act update
Related Regulations
- Thailand Electronic Transactions Act: this is the broader legal foundation for digital transactions and ETDA’s regulatory role, and it likely interacts with AI governance where AI services are delivered electronically.ETDA homepage, Thailand regulatory landscape PDF
- Thailand Personal Data Protection Act (PDPA): AI systems processing personal data must comply with Thai privacy requirements, especially for training, inference, profiling, and automated decision-making workflows.Thailand PDPA summary, Thailand regulatory landscape PDF
- NIST AI Risk Management Framework: not a Thai regulation, but a useful governance framework for mapping risk, trustworthiness, and lifecycle controls to AI obligations.NIST AI RMF, NIST AI RMF playbook
- ISO/IEC 42001: an AI management system standard that can support policy, controls, internal audits, and continuous improvement for organizations preparing for AI regulation.ISO/IEC 42001 overview, ISO AI management systems
- OECD AI Principles: these principles are frequently used as a benchmark for trustworthy AI and may inform Thai governance expectations.OECD AI Principles, OECD AI policy observatory
Industry Impact
- Financial services: lenders, insurers, and fintech firms will likely feel the strongest impact because they often use high-risk scoring, fraud detection, and automated decision systems that may fall into regulated categories.Thailand regulatory landscape PDF, Thailand tech governance analysis
- Healthcare: AI used in diagnosis support, triage, or clinical workflow automation will likely face stronger documentation, validation, and incident-management expectations if classified as high-risk.Thailand tech governance analysis, Thailand draft AI act commentary
- Consumer tech and platforms: chatbots, generative-AI products, and social or media platforms may need new disclosure tooling for AI-generated content and deepfakes.Thailand tech governance analysis, Thailand draft AI act commentary
- Government and critical infrastructure vendors: if the draft retains special obligations for government agencies and critical infrastructure, vendors may need stronger contractual, hosting, and localization controls.Thailand draft AI act commentary, Thailand AI regulation overview
- Market-entry and procurement effects: organizations selling into Thailand may face longer review cycles, higher compliance costs, and more formal procurement due diligence for AI-enabled products.Thailand draft AI act commentary, Thailand business reaction report
Sources
- [Thailand draft AI act commentary](https://www.bakermckenzie.com/en/insight/publications/2026/07/thailand-draft-ai-act
Recent developments
- — Analysis of Thailand’s Draft Artificial Intelligence Act highlights a risk-based framework centered on a new National AI Committee, with the draft out for public consultation from 9 July through 14 August 2026 and positioned as a cornerstone of Thailand’s emerging tech-governance regime.[11] (source)
- — Legal briefing identifies seven key implications of the Draft AI Act released for public consultation on 9 July 2026, including prohibited and high‑risk AI categories, strict and joint liability, powers to suspend or block AI systems, and urges stakeholders to submit comments before the 14 August 2026 deadline.[1] (source)
- — News report describes the Digital Economy Minister’s push to finalize the landmark AI Act during the current fiscal year, noting concerns from some business and investment stakeholders about potential adverse impacts while the government frames the law as vital for responsible AI growth.[8] (source)
- — Law firm commentary explains that Thailand’s proposed AI law adopts a risk‑based regulatory framework inspired by the EU AI Act, with the draft open for public consultation until 14 August 2026 and expected to rely heavily on delegated legislation and sectoral regulators.[4][9] (source)
- — Business‑focused analysis outlines what organizations should prepare for under the Draft AI Act released 2 July 2026, emphasizing obligations for AI deployers (not only developers), risk‑management systems, log‑retention, reporting of unforeseen risks, and potential administrative fines of THB 1–5 million, while stressing the text is still only a draft.[5] (source)
- — Detailed client alert notes that on 9 July 2026 ETDA issued the Draft AI Act for public consultation, categorizing AI into prohibited, high‑risk, licensing, and transparency‑obligated systems (including deepfakes, chatbots, and generative AI), with strong enforcement tools and significant compliance impacts for AI providers and deployers.[1] (source)
- — Practitioner update reports that Thailand began consulting on an updated Draft AI Act on 2 July 2026, summarizing key features such as a tiered risk‑based system, extraterritorial reach, transparency requirements for AI‑generated content, potential data‑localization for important AI services, and a strict liability regime for AI‑related damages, noting the law is not yet in force.[6] (source)
- — Policy and regulatory briefing describes the July 2, 2026 draft AI Act as combining EU‑style risk tiers with no‑fault strict liability, mandatory local representatives for foreign providers, AI‑content labeling, and significant fines, and clarifies that the text is under public hearing for roughly 30 days and has not yet reached Cabinet or Parliament.[3] (source)
- — Specialist regulatory news analysis explains that Thailand’s first draft AI bill blends an EU‑inspired, risk‑based framework with Thai regulatory practice, combining strict civil liability, regulatory sandboxes, standards and certification, and broad powers for sector‑specific regulators, while leaving many detailed obligations to future royal decrees and guidance.[9][10] (source)
- — Client alert announces that ETDA released a new version of the Draft Artificial Intelligence Act on 2 July 2026 for about a 30‑day public hearing, highlighting extraterritorial scope, tiered risk classification, strict liability for AI‑related damage, transparency requirements for AI‑generated content, and phased implementation timelines affecting global AI developers and deployers.[2] (source)
Related regulations
- Brazilian Artificial Intelligence Act — Brazil, Proposed
- NIST AI Risk Management Framework (AI RMF 1.0) — United States, Active, effective 2023-01-26
- Artificial Intelligence and Data Act — Canada, Superseded
- Colorado Artificial Intelligence Act (SB 24-205) — Colorado, Superseded, effective 2026-06-30
- Texas Responsible Artificial Intelligence Governance Act (TRAIGA, HB 149) — Texas, Active, effective 2026-01-01
- EU AI Act - Annex III High-Risk System Requirements (2 Dec 2027) — European Union, Upcoming, effective 2027-12-02
- EU AI Act - GPAI Model Obligations (2 Aug 2025) and Enforcement (2 Aug 2026) — European Union, Active, effective 2025-08-02
- EU Digital Omnibus on AI (Regulation (EU) 2026/1744) — European Union, Active, effective 2026-07-27
Put it into practice
- Generate the policy: NIST AI RMF policy generator (generatepolicy.com)
- Buy the policy pack: NIST AI RMF Implementation Policy (cyberpolicy.shop)
- Build it yourself: Pillar 06 Companion — The 2026 AI Risk Register (ciso.diy)
Browse all regulations · Compliance deadlines · Latest updates