On 2 July 2026, Thailand's Electronic Transactions Development Agency (ETDA) released a revised Draft AI Act for public consultation, with the consultation period closing on 14 August 2026. If enacted, the Draft AI Act would be Thailand's first statute imposing registration, conformity assessment, incident reporting, and liability obligations on developers, providers, deployers, and platform operators of artificial intelligence systems.
The draft establishes four risk tiers: prohibited AI systems, high-risk AI systems, AI systems requiring an operating licence, and AI systems subject to transparency obligations. Transparency obligations apply to systems generating deepfakes, operating as chatbots, or producing generative AI output. High-risk AI systems must undergo conformity assessment before deployment, register with the ETDA, implement post-market monitoring, and report incidents, mirroring obligations in Articles 9 to 17 of the EU AI Act. The draft departs materially from the EU AI Act on liability: it imposes strict liability for AI-related damage without requiring proof of fault on the part of the developer, provider, or deployer.
Foreign developers, providers, and deployers whose AI systems are used in Thailand fall within scope regardless of where their servers or registered offices are located. Those without a registered Thai presence must appoint a local representative to act as point of contact with the ETDA, a requirement analogous to Article 25 of the EU AI Act. Businesses affected include providers of AI-enabled financial services, insurance underwriting, medical diagnostics, vehicle control systems, critical infrastructure management, and consumer platforms directing services at Thai users.
The strict liability provision is the draft's most contested element. Industry groups have raised concerns during the consultation that unlimited strict liability will deter deployment of beneficial AI applications in Thailand, and the ETDA has indicated the provision may be revised before finalisation. The ETDA intends to complete the legislative process before the end of 2026, though parliamentary scheduling and committee review timelines remain unconfirmed. The draft may change materially following the consultation.
Licentium advises technology companies, financial institutions, and multinational businesses on AI regulatory compliance across EU, UK, and Southeast Asian jurisdictions. Work we undertake includes AI system risk-tier classification, high-risk AI compliance programme design, local representative appointment, cross-border AI deployment structuring, and multi-jurisdictional AI regulatory horizon scanning.