The gun was legal

On 7 August 2026 a school in Nonthaburi was attacked. The weapon, according to police, belonged to the attacker’s grandfather and was lawfully registered to him. It was taken from the house, along with ammunition, by a fourteen-year-old.

I am not going to describe what happened, or who did it, and there will be no figures in this article about how many people were hurt. None of that is what I can usefully add, and the analysis below does not need it.

What I can add is where the gun came from, because that is the part almost nobody is discussing and it is the part that decides what any new law should do.

The number everyone uses

The Small Arms Survey’s Civilian Firearms Holdings table puts Thailand at 10.34 million firearms in civilian hands — around 15.1 per 100 people, the highest rate in Southeast Asia and thirteenth in the world by total number. Cambodia is estimated at 4.5 per 100, the Philippines 3.6, Malaysia 0.7.

Read the label before using the number. That table is dated 2017. It is calculated against a 2017 population of 68.3 million, and Thailand’s row is marked as an expert estimate rather than a survey or a count — the Survey’s own middle category of confidence. It is the best figure anybody has and it is nine years old.

Within that estimate, 6.22 million are registered and roughly 4.12 million are not — so nearly forty per cent of the modelled total sits outside the register. The unregistered half of any such estimate is the part built on the least evidence, which is worth holding on to.

Four million illegal guns is the figure that leads every article about this, including most of the Thai ones, and it is real. In the 2023 fiscal year Thai authorities recorded 72,110 firearm and explosive cases, of which 34,806 concerned illegal weapons.

But look at what the last two school incidents actually used.

August 2026, Nonthaburi: a registered firearm, lawfully held, taken from a family home.

October 2023, Siam Paragon: a blank-firing gun — a legal object, sold openly — that had been modified. It was bought online.

Neither of those is a gun that was smuggled in. One was inside the registration system. The other was inside a category the law does not treat as a firearm at all. A policy aimed exclusively at the four million reaches neither route. Whether it would have changed either outcome is not something I can tell you, and I am not going to imply it.

What the licence actually asks

Here is the thing I could not find explained anywhere in English, and it is the centre of this.

A Thai firearms licence — ป.4, issued by the Department of Provincial Administration — is granted to a named person for a stated purpose: self-protection, protection of property, sport, or collection. Applicants must meet conditions on age, residence, occupation and record.

That is the point at which the state looks hardest at you, and under the present arrangement it is very close to the only point.

Then, as far as the instruments themselves go, nothing.

The Department of Provincial Administration publishes a service guide for the ป.4. It lists the documents, the ฿100 and ฿500 fees, the fifteen-day processing time and the steps the registrar takes. It sets out no expiry date, no renewal, and no requirement ever to present the holder or the weapon again. Nothing in the Act’s own structure supplies one either.

I need to be careful here, because this is the load-bearing claim in the article and the draft of this piece had it wrong. A five-to-ten-year re-presentation cycle circulates in Thai secondary sources and the draft repeated it. It is not in the Act, not in the department’s published guidance, and I could not find the instrument that would impose it. The decisive evidence is what the government proposed in August 2026, which is to introduce a renewal period. You do not introduce renewal for a licence that already renews.

So: no annual return, no standing requirement to demonstrate how the weapon is stored, no periodic re-examination of the holder’s circumstances, and no published ceiling on how many licences one person may accumulate — although the application form does ask for copies of every ป.4 already held, which means the state can count them and has chosen not to cap them.

A licence granted in 1998 to a person whose circumstances no longer resemble their application is, in the ordinary case, still a licence.

Which is article 81’s finding again, on incomparably higher stakes. That piece found that Thai traffic fines went overwhelmingly unpaid — not through corruption, but because two government databases were not connected and the consequence had nothing to attach to. The rule was real. The mechanism was missing.

The firearms regime has a strong front door and almost nothing behind it. The state’s attention is concentrated at issue, and issue is the moment when a firearm is least likely to be a problem.

The state was also the wholesaler

Now the part that has never appeared in English, and it is the reason “gun culture” is the wrong frame.

โครงการปืนสวัสดิการ — the welfare firearms scheme — allows certain state employees to buy guns at below-market prices through their own agency. Civil servants, police, state enterprise staff and, in some schemes, lawyers. The stated purpose is self-defence, protection of property, and duty-related security work.

This is not an informal arrangement, and you do not have to take my word for it. The department’s own published guide to obtaining a ป.4 lists purchase from a project procuring firearms as a welfare benefit for officials alongside a gun shop and a private transfer, as one of the three ordinary ways a person arrives at a firearms registration.

The state is not only the licensing authority. In this channel it is the supplier, and the price is a staff benefit.

Two features of that scheme are doing more work than anyone intended.

First, the guns can leave. A Ministry of Interior instruction — มท 0307.4/ว 1479 — permits a welfare firearm to be transferred into another person’s name after five years of possession, inheritance aside. Published Thai legal research on the Firearms Act identifies this as a route by which subsidised weapons reach the general population, and the same research notes that scheme weapons have been found pawned or otherwise held contrary to the scheme’s purpose.

Second, there is no clear cap. The research finds no unambiguous limit on how many welfare firearms an eligible official may purchase.

So there is a lawful pipeline running from a state-subsidised price to private hands, on a five-year timer.

The police version of the scheme was suspended indefinitely in October 2022, following a case in which firearms were reported taken from a station armoury and sold on. I want to be careful with that one — it is a single reported case and I have not seen the disposal of it — but the suspension is a matter of public record and a national police force does not suspend its own welfare programme over nothing.

And in August 2026 the pause became much wider. Among the measures ordered on 11 August, the approval of new welfare firearms projects was halted, along with the issue of ป.3 purchase permits, both pending a Cabinet decision and both described as temporary. These are administrative suspensions, not changes to the law. The provision that lets a welfare weapon change hands after five years is untouched by them, and so are the licences already issued.

This is article 89’s finding wearing an unfamiliar coat. The site has argued repeatedly that Thai systems fail when access is categorical rather than universal, because the category is drawn around people who already had options. Here the category is state employment and what it grants is not a protection. It is a discount on a firearm, and after five years the discount walks out into general circulation with the gun.

The category the law does not see

The third gap is definitional, and the Paragon case exposed it.

A blank-firing gun is not, in Thai law, a firearm. It is a สิ่งเทียมอาวุธปืน — an imitation — and the Act gives it a short chapter of its own, sections 52 to 54.

The gap is written into the definition. An imitation is สิ่งซึ่งมีรูปและลักษณะอันน่าจะทำให้หลงเชื่อว่าเป็นอาวุธปืน — a thing whose form and characteristics are likely to cause the belief that it is a firearm. The statutory test is resemblance. It is not capability. The object is classified by what it looks like, and some of these can be modified to fire live ammunition.

One correction to how this is usually put, including in the draft of this article. It is not true that imitations are unlicensed. Section 52 requires a licence from the local registrar to order, import or trade them. What needs no licence is buying or holding one. The Act controls the channel and leaves the object alone — which is the precise shape that an online marketplace listing is well suited to slip through.

So the legal question “is this a firearm?” and the practical question “can this kill someone?” are answered by different documents.

This is a familiar shape on this site. Article 88 The sticker is not insurance found that Thailand’s compulsory motor sticker is not the insurance most people believe it to be, because the object and the protection are two different things that look like one. Here an object that is legally not a gun is, after an afternoon’s work, a gun.

What is being proposed

The legislative response is moving quickly and is worth stating plainly, because a reader in September will want to know what actually changed.

Two things are in force, and both are administrative rather than legal. Under measures ordered on 11 August 2026, the issue of ป.3 purchase permits was suspended and the approval of new welfare firearms projects was halted, both pending a Cabinet decision and both described as temporary. Alongside them came a nationwide audit of ป.3 permits, a drive to regularise ป.4 registrations, inspections of dealers and ammunition stocks, and hotline 1567 for reports of illegal weapons and public carrying.

Everything else is a proposal, and none of it is law. The government asked for an amended Firearms Act within sixty days — a deadline falling in October 2026, after this article publishes. The Council of State opened a public consultation on the principles of a new Act on 12 August 2026. A party-sponsored bill of around a hundred sections has been published separately. Between them the measures under discussion include a fixed renewal period for the ป.4 — three years is the government’s figure, five years appears in the draft bill — along with physical and mental health checks, a ceiling on how many firearms one person may hold, a firearm-identity database, wider powers for the registrar, and an amnesty for surrendering unregistered weapons.

A government spokesperson said on 12 August that drafting was still under way and that conclusions about what the law now requires were premature. That was still the position when this article was checked on 4 September 2026: the suspensions are real, and the new Act is not yet written.

Then the vehicle changed. On 8 September 2026 the Cabinet approved in principle a draft emergency decree amending the 1947 Act, rather than waiting for a new one. As the Cabinet’s spokesperson described it, the decree would add training, a test and a mental-health check to the licensing process, shorten the term of a licence and end the licence for life, and open a ninety-day amnesty for surrendering unregistered weapons to the registrar without penalty. Approval in principle is not enactment. An emergency decree takes effect when it is published in the Royal Gazette, and on 13 September the government’s spokesperson said the change was still at the draft stage. When this article was re-checked on 27 September 2026, no publication had been reported. If it is published, the licence this article describes, one that never expires, ends with it.

Renewal is the one to watch. Everything above about the front door and the empty corridor behind it is a description of a licence that does not expire. A licence with a renewal date is a mechanism, and this site’s consistent finding is that mechanisms are what Thai rules are usually missing — not severity.

The amnesty is the one to be sceptical about, and I say that as someone who thinks it is probably still worth doing. It will collect weapons from people who were never the risk. That is not nothing — a surrendered weapon is one fewer in circulation — but it should not be counted as having addressed the four million.

What I cannot tell you

Several things, and the gaps matter more than usual here.

I do not know how many welfare firearms have been issued, in total or per year. I could not find a published figure and I am not going to estimate one.

I do not know how many have been transferred out under the five-year provision. That number would settle whether the pipeline is a technicality or a significant channel, and it is the single most useful statistic nobody appears to have published.

Nobody knows the real number, and neither do I. The 10.34 million is a 2017 expert estimate, and its unregistered component is by its nature the least knowable part of it. It is the best figure available and it is not a census. Nine years on, nothing has replaced it.

And I do not know whether any of this would have prevented any particular incident, which is the question everyone actually wants answered and the one that cannot honestly be answered by anybody.

Common misconceptions

“It’s a black market problem.” An estimated 4.12 million weapons are outside the register, and the two most-discussed school incidents used a legally registered gun and a legally sold blank gun respectively.

“Guns are hard to get legally in Thailand.” The licence has real conditions at application. Thailand still has the highest estimated civilian ownership rate in Southeast Asia.

“A licence is checked regularly.” The department’s published guidance for the ป.4 sets out no expiry, no renewal and no repeat presentation. That a three-year renewal was proposed in August 2026 is the clearest available evidence that there is not one now.

“Blank guns are toys.” The Act classes them as imitations rather than firearms, on a test of resemblance rather than capability. Trading them requires a licence; buying one does not. Some can be modified to fire live rounds.

“The state only regulates guns.” Through ปืนสวัสดิการ it has also sold them, at below-market prices, to its own employees.