The proposed TH-AI Passport project has become one of the most debated digital policy initiatives in recent years. Framed as a national investment in human capital and artificial intelligence capability, the scheme has drawn scrutiny over its scale, procurement process and value for money.
At its core is a 1.6-billion-baht plan to provide access to advanced AI tools for up to 5 million people or roughly 10% of the eligible population aged 15 and above, while also generating data and skills intended to strengthen the domestic AI ecosystem.
What is the TH-AI Passport project?
The TH-AI Passport is initiated by Ministry of Digital Economy and Society (DES) to provide Thai citizens with access to multiple AI models (around 12 platforms and 24-25 models) through a single subscription framework.
DES Minister Chaichanok Chidchob says the scheme is funded through the Digital Economy and Society Development Fund (DE Fund) rather than the core state budget, citing urgency as the reason for its implementation this month, ahead of the next fiscal budget cycle in October. It targets students, workers, state officials and the public aged 15 and above, with particular emphasis on first-job entrants entering the labour market.
He describes the programme as both a digital access initiative and a skills development platform, linking AI usage with structured training under a "Learn to Earn" framework.
Why is the project necessary?
Mr Chaichanok said the project responds to weak national AI adoption.
Thailand's AI diffusion rate is placed at 10.7%, compared with a global benchmark of 16.3%, based on Microsoft AI Economy Institute–linked adoption estimates. Regional peers such as Vietnam (23.5%) and Singapore (60.9%) are significantly ahead. He frames this gap as a competitiveness risk, saying AI is a general-purpose technology with economic effects comparable to electricity or the internet.
The policy is also linked to labour-market concerns, with AI positioned as a tool to offset productivity constraints linked to demographic changes.
Who stands to benefit?
The government says the scheme will give users access to premium AI tools at a heavily reduced cost, estimated at around 27 baht per month through bulk procurement, which users do not need to pay as it will be subsidised.
Students, workers, public servants and SMEs are expected to benefit from improved productivity and access to tools that are normally subscription-based.
Beyond individual users, the project is also tied to national AI development efforts, particularly the Thai Large Language Model (ThaiLLM), an open-source, open-licence AI infrastructure tailored specifically for the Thai language.
ThaiLLM is a collaborative effort between the Big Data Institute and leading institutions including the National Electronics and Computer Technology Centre, Mahidol and Chulalongkorn universities. The ThaiLLM is designed to improve AI understanding of Thai language structure, culture and documentation, reducing reliance on foreign models.
Critics, however, argue the link between mass subscriptions and meaningful domestic AI development remains unproven.
How could the project be expanded?
Expansion is expected to focus on increasing user coverage, improving AI training integration and strengthening domestic AI infrastructure.
It also aligns with broader plans to develop the AI ecosystem, including research into multimodal systems and more advanced "Agentic AI", where multiple specialised AI systems work together under a coordinating framework.
Longer-term objectives include expanding ThaiLLM capabilities and increasing its integration with public and private sector applications.
What safeguards exist for data privacy and security?
DES permanent secretary Patchara Anantasilp said the system is governed under national data protection rules, with anonymisation applied before analysis and separation between identity verification and usage data.
AI providers are prohibited from using personal prompts for model training, and data is stored on Thailand-based cloud infrastructure.
However, key governance details remain unclear, including audit mechanisms, retention periods and independent verification of anonymisation standards.
Critics said without transparent enforcement structures, assurances on data protection remain largely procedural rather than independently tested.
Why was this procurement model chosen?
Mr Chaichanok said a consortium model allows integration of multiple AI systems under a single access platform, while also enabling bulk pricing and centralised management.
But Pawoot Pongvitayapanu, People's Party MP, has questioned why the state did not pursue direct licensing agreements with major AI providers.
He said such arrangements are already used in other countries and could reduce administrative complexity while improving transparency and price comparability.
Instead, he warned the structure introduces additional layers between public funds and end providers, making it harder to independently verify cost efficiency and contract value.
Are procurement concerns justified?
Opposition MPs, including Rukchanok Sirnok of the People's Party, have raised questions about whether projected demand for 5 million users reflects real usage patterns, or whether it is primarily a funding justification.
She has also pointed to the speed of the tender process -- 34 days -- as unusually short for a multi-billion-baht digital infrastructure project, even if it meets minimum legal thresholds.
Mr Chaichanok said the tender followed standard procedures. A public announcement was made on Nov 17 with an open bidding period lasting 34 days, exceeding the legal minimum of 20 days.
The minister also notes the process included public consultation, with around seven interested parties during the pre-bid phase and three consortiums submitting proposals.
Further scrutiny has focused on contractor concentration risks in determining pricing benchmarks for multi-billion-baht projects and whether procurement specifications may indirectly favour certain business arrangements.
The project has been called in parliament for further review by oversight bodies, including the National Anti-Corruption Commission and the State Audit Office.
Is the project worth B1.6bn baht?
The government describes the scheme as digital infrastructure investment intended to accelerate AI adoption from 10.7% towards regional benchmarks and improve workforce productivity.
But opposition figures argue the project raises unresolved questions about procurement design, demand assumptions and value for money.
The central tension is whether the scheme is primarily an efficiency-driven national upgrade, or a complex procurement structure built around contested demand projections and limited transparency.
With parliamentary scrutiny and oversight agencies expected to examine elements of the tender process, the TH-AI Passport is likely to remain under political scrutiny well beyond its planned rollout.
For now, the project sits in a narrow space between policy ambition and institutional doubt, with its final judgement likely to depend less on rollout numbers than on whether its underlying design can withstand scrutiny.
