Following Thailand’s 2026 general elections, the Election Commission of Thailand (ECT) has faced a legitimacy crisis over the transparency of the polls with over 5,000 irregularities being reported.

Thailand held elections on Sunday (8 February), with only 65 per cent of eligible voters turning to cast their votes. The figure was considered the lowest in three decades and down from 75 per cent in the 2023 general election.

Prinya Thaewanarumitkul, a law professor who has followed Thailand’s elections since 2005, posted on Facebook his observation that the turnout figure was the most abnormal issue in this year’s election.

At the same time, abnormalities were flagged just shortly after the polls closed on Sunday, raising concerns over electoral fraud. All eyes have been on the ECT, which was allocated over 7.8 billion of taxpayers’ money to stage the election.

All polling stations across the country finished vote counting late on election night, but as of Saturday (14 February), the ECT’s website revealed only 95 per cent of votes recorded, raising concerns that the record would be later adjusted or altered.

On Thursday (12 February), the ECT urged the general public and the media not to use the recorded votes on the ECT’s website as a reference on the grounds that the results have yet to be made official.

The unofficial results from the ECT also significantly differed from those recorded by Vote62, an independent election watchdog, which relied on over 16,000 election volunteers in 77 provinces across the country who recorded votes from the polling stations.

Since election day, over 5,000 irregularities, ranging from missing candidate and voter information to valid votes being counted as invalid, have been reported to Vote62.

Discrepancies between the figures recorded in the ECT’s online system and those in polling stations have been spotted in several constituencies across the country. 24 constituencies in eight provinces have sought recounts.

In Chonburi, for example, hundreds of protesters have been gathering at a badminton court in Mueang Chonburi district, where the ballot boxes are being kept, since last night (9 February).

They called for a recount after some abnormalities were reported. In some polling stations, there were more votes than the number of ballots. In others, polling officials continued to count votes despite a power outage, raising concerns about fraud because observers could not see the ballots as they were being counted. At the badminton court, it was also found that ballot boxes were not sealed properly, and ballot log sheets were found in the trash.

However, the ECT decided not to order a recount, concluding that there was no solid evidence to prove that the election was fraudulent. The Bhumjaithai candidate for the constituency, Suchart Chomklin, Caretaker Deputy Prime Minister, is leading the People’s Party candidate Warot Sirirak by 3,783 votes.

The case drew public attention and created a ripple effect, sparking further controversies and increasing pressure on the ECT.

In several areas, the number of recorded votes significantly outnumbered that of eligible voters who turned out.

The situation has sparked outcry and protests to demand recounting all votes nationwide. The hashtags “Nationwide Recount” and “ECT scams elections” have been trending on Thailand’s X platform since 9 February.

In some areas, vote counting revealed while one party was recorded as receiving not even a single vote, people in the constituency insisted that they themselves cast votes for the party, demanding explanations for the missing votes.

In the elections, the number of constituency MPs and party list MPs’ ballot papers must match, as voters were given the two ballot papers simultaneously before entering the polling booths.

However, discrepancies between these two ballots were detected across the country. According to Vote62, only 10 out of 400 constituencies have no such issue. In constituency 5 in Roi-Et province, for example, the discrepancy was over 8,000. These inconsistencies have added fuel to public outrage.

The ECT stated that the figures were unofficial 95 per cent results. Once counting reaches 100 per cent, the figures will be fixed.

Sarinee Achavanuntakul, an independent academic, argued that despite the claim, the figures remain irrational. The current cumulative discrepancy between constituency ballots and party-list ballots on the ECT’s website exceeds 66,000, and that it would be mathematically impossible to reduce the discrepancy to zero whenr the remaining five per cent of ballots are counted. She noted that such claims implied that the figures may eventually be altered.

Delayed reporting amid transparency concerns

Yingcheep Atchanont, Director of iLaw, explained that once the vote counting was completed, the officials at the polling stations would upload the results to the ECT’s back-end system.

He observed that during the uploading process and the reporting of the results on ECTScore69, there is a follow-up stage, which allows certain individuals to access the system, raising concerns that this could allow an intermediary to intervene in the vote data before it is finalised.

At around 7.00 p.m. on election day, the reported results stalled at 4 per cent for an extended period, before it froze again at 18% for nearly an hour between 7:30 and 8:25 pm. The ECT later stated that the system had crashed and could not process the data.

Yingcheep remarked that while vote processing was not complete, Bhumjaithai appeared to secure the most votes with 200 seats at 10.30 p.m., leaving behind the People’s Party, which at the time had secured only 70 seats. He noted that the gap between the first and second-ranked parties was unusually wide.

At the time (10.30 pm), in the provinces where Bhumjaithai secured the most seats, it was found that the percentage of votes displayed on the website was significantly higher when compared to the provinces where the People’s Party was taking the lead.

He observed that there were intermediaries to handle votes before they entered the ECT’s public reporting system, as the figures available to the general public were not reported in real time.

In addition, some citizens also noticed that, as reported, the number of votes for certain political parties was adjusted downward in some areas. Yingcheep said that reports of the vote numbers should only increase as more votes are counted.

He reiterated that vote counting was completed on the election day, and all votes were submitted to the ECT’s back-end system, raising questions about why the remaining five per cent have yet to be released. Meanwhile, the ECT stated that the figures relied on officials from the polling stations, and it has yet to be official, noting that there may have been some human errors.

Secret ballot may not be secret

On Thursday (12 February), another issue was raised, and it could potentially lead to a void election. Barcodes were found on the ballot papers for both constituency MPs and party list MPs. When scanned, they yielded different codes for each ballot paper, with one letter and eight digits.

It was found that each ballot paper could be scanned and connected to the ballot stub which voters were required to sign. The ballot stubs also contained voters numbers as given in the lists of eligible voters in each constituency. This makes it possible to identify which voter completed each ballot paper and which MPs or parties they voted for.

According to the Constitution, MPs shall be elected by “secret ballot,” and no identifying marks shall be placed on ballot papers. Otherwise, the elections could potentially be declared void.

On Friday (13 February), the ECT held a press conference regarding the barcodes and QR codes, citing the provision which allows it to prescribe codes or marks on ballot papers as a safeguard against “forgery”.

The barcodes were also used to track the polling stations where the ballots were distributed, and that would help prevent ghost voting and electoral fraud.

The ECT also asserted that tracking the identity of individual voters is impossible. It said that doing so would require access to additional elements, including ballot stubs and voter name lists, which are now kept separately, and that the Commission could not verify whether the numbers on the barcodes align with the ballot stubs as requested by the media and general public.

When asked on which occasions the ECT would scan barcodes, it said it would only do so if a complaint was filed, as the ECT must identify the polling station where an issue arose. The ECT admitted that the barcodes could trace back to the ballot stubs, reiterating that they were included for investigation purposes and to maintain transparency.