Korn Chatikavanij, Deputy Leader of the Democrat Party, highlights five issues with the 1.6 billion baht TH-AI Passport project potentially wasting funds. He calls for a review and suggests shifting investment toward building Thailand's economic foundation to support SMEs in AI technology sectors.
Mr. Korn Chatikavanij, party-list MP and Deputy Leader of the Democrat Party, posted the following on his Facebook page, Korn Chatikavanij - Korn Chatikavanij.
This will be a bit long, but let's understand together another current issue.
In the past few days, the TH-AI Passport project worth over 1.6 billion baht became a hot topic after Mr. Pawut Pongwitthayaphanu from the People’s Party raised a live question in parliament, asking Minister of Digital Economy and Society Chaiyachon Chidchob about suspicious details.
I was there listening myself. Few MPs were in the meeting room, likely because most had dispersed to attend committee meetings elsewhere.
When Mr. Pawut began asking the first question, I immediately texted the Democrat Party group, saying, “There’s a live question on the issue we’re examining. Pawut is asking well, and Chaiyachon is about to respond.”
The more I follow this, the more problems I see with the project. Society should question whether the budget was spent efficiently and if better alternatives exist.
Next, the opposition invited the Ministry of Digital Economy this week to explain to the Anti-Money Laundering Committee, chaired by Mr. Phitakdech Deddecho, and we coordinated for Pawut to join the discussion.
I consulted with several insiders to raise questions and better understand the matter.
1. What has the country gained from spending billions?
How many people truly benefit? The government claims it will grant free one-year Pro AI access to 5 million Thais, funded by the DE Fund at about 324 baht per person per year. Yet basic AI functions are already freely accessible to most. Interestingly, the figure of 5 million seems not based on real demand but rather on knowing that about 1.6 billion baht remained in the fund, then designing a project to match that amount.
Importantly, the project shows major contradictions in actual user behavior. Globally, public platforms already offer free access. The free version is sufficient for daily tasks like translation or document summarization.
Most people lack AI literacy skills, not Pro access. Distributing Pro licenses widely without building skills is ineffective. For developers or builders pushing innovation, the government’s Pro monthly package is inadequate due to hourly command limits.
My team explained that serious users need API access charged by token consumption. The government is spending over a billion baht buying Pro licenses to distribute broadly to those who don’t need them, while not supporting advanced systems for heavy users. Of the 5 million licenses, few will create real economic value. Moreover, the subscription model means unused licenses expire after one year; funds are fully spent, and unused rights become pure profit for contractors. The government cannot reclaim or roll over these funds.
I think this final fact is very concerning.
My basic principle is that when spending billions, the government must measure not just how many claim rights, but what additional economic value Thailand gains from this investment.
2. We are paying annual rent instead of building assets.
By law, the DE Fund should develop core and long-term digital infrastructure. But this project functions as operating expenses, paying a large sum annually to rent AI services. After one year, the 1.6 billion baht is gone along with user rights. Approval was made by the fund board without the normal parliamentary budget review process. The government’s claim that this is cost-effective compared to investing in personnel development misses the point. Buying ready-made rights for users only creates users, not technology creators like builders. At contract end, Thailand walks away empty-handed with no tech assets or innovations.
3. Having servers in Thailand does not mean we own the data.
This is the most confusing point. The government says the project runs in a data center located in Thailand, with user data and prompts stored on cloud servers inside the country. AI providers access data anonymously, ensuring data sovereignty. But we must distinguish between physical location and technology ownership.
Even if foreign companies build data centers in Thailand, the software, AI models, and behavioral data repositories that generate value and profit remain 100% foreign-owned. So with a rental contract, even if servers are here, when the lease ends, the system can be shut down, and money and technology value flow out of the country.
4. Transparency issues in the TOR.
My team explained that the project’s tender documents are very vague. The largest expense of 1.5 billion baht is broadly described as AI tokens costs, with no clear brand breakdown, preventing backend system audits. It also requires supporting 500,000 simultaneous users per hour, but actual calculations show support for only 139 users per second, raising doubts about capacity.
This excludes the rushed bidding period of just 34 days during the New Year holiday, and the Minister of Digital Economy’s attitude, saying he is not concerned about who gets the contract. As the highest official, he cannot shirk responsibility for tax money oversight and ensuring legal transparency.
5. I asked my team if there are more sustainable alternatives to spending this 1.6 billion baht rather than just renting rights for one year without data ownership.
The answer is yes. Instead of yearly rental payments to foreigners, the government should reallocate funds to build more valuable foundations. For example,
- Convert this sum into investment for high-performance computing capacity, such as GPU clusters owned by Thailand, housed in private data centers within the country. These tools last 3 to 5 years, becoming national assets accessible to startups and developers without recurring rental fees, enabling true shared ownership in the innovation value chain.
Or promote Thailand as an AI Builder Hub, a proactive national strategy proposed by the Democrat Party. This shifts focus from chasing technology to establishing an innovation ecosystem. For instance, repurposing government spaces like the Tobacco Factory 5 building into infrastructure linked to state-owned processing centers to create sandbox environments where startups and small teams (even just two people) can experiment and build AI products that generate revenue. Coupled with public procurement mechanisms attracting financial institutions, funding sources, and global tech companies to invest jointly, this enables technology transfer, high-level employment, and elevates Thailand as a genuine innovation hub.
Alternatively, transform the budget into innovation vouchers co-invested by government and private sector so SMEs can hire experts or purchase AI systems tailored to their specific needs.
Or invest in developing Thai language models and intelligent data repositories in collaboration with universities on niche or specialized topics currently hindered by data access challenges, building long-term intellectual property assets for the country.
It’s good for Thais to use AI, but the government must reconsider what benefits this billion-baht budget brings to the Thai economy, who will truly use it to create value, and whether spending money to enrich foreigners through unused rental contracts or building AI infrastructure truly owned by Thais offers better value for taxpayers’ money.