RN
Twenty-five thousand domains, and no published definition of what put a domain on the list.
That is the fact worth keeping from Nepal's 2018 blocking order. The figure came from the telecommunications authority itself as the campaign expanded through the autumn, and a list that large is not compiled by review — it is compiled by keyword, by category feed, or by whatever a filtering vendor supplies. Which is fine as engineering and fatal as law, because the instruction it implements never said what counted.
What actually happened
The communications ministry issued a statement in late September 2018; the telecommunications authority passed a directive to internet service providers; the providers began blocking. Non-compliance carried a fine and the risk of an operating licence, so compliance was not really optional.
The legal hook was the country's criminal code, then newly in force, which prohibits producing, publishing and distributing obscene material with penalties of up to a year. The important thing about that provision is what it is aimed at — production and distribution, both of which were already offences before anyone blocked anything. The order added no new prohibition. It added an access restriction, applied to consumption, on the authority of a rule about supply.
And it defined nothing. No test for obscenity, no criteria, no published list, no appeal route for a site that thought it had been swept up. Media-freedom groups said so at the time and the objection was structural rather than partisan: a blocking power that does not say what it blocks is a blocking power that can be pointed anywhere later.
The reasoning that was offered
The order followed a period of protest over the handling of a serious criminal case, with public anger directed largely at the police and the authorities rather than at any technology. The government's stated rationale connected easy access to explicit material to sexual violence.
That connection has not been established by research, and this article does not assert it in either direction. What can be said is that the measure was introduced as a response to failures of investigation and prosecution, and that it did nothing to address them — which is precisely what Nepali critics said at the time. The most consistent objection from women's-rights and internet-freedom voices was not about censorship in the abstract but about substitution: attention and political credit were being spent on a technical gesture instead of on the justice system that had produced the anger.
Sex educators added a narrower point that has aged well. An undefined obscenity filter does not distinguish between explicit material and sexual health information, and in a country where formal sex education was thin, the second was collateral.
Why it could not work as stated
The service providers said so publicly, and they were the ones being ordered to do it.
Blocking is per-domain and per-provider. New domains appear faster than any list can be maintained; a site blocked on one network stays reachable on another; a VPN or a different resolver removes the obstacle entirely. None of this is obscure, and none of it was disputed — an authority official's own figure of tens of thousands of blocked domains is itself a description of a treadmill rather than an achievement.
The result is the familiar asymmetry. Anyone motivated enough to look for a workaround finds one in minutes. Everyone else is filtered — including the people looking for health information they cannot get elsewhere.
What it set up
The 2018 order established, in practice, three things: that the executive could instruct ISPs to block categories of content without legislation; that the category need not be defined; and that providers would comply because their licences depended on it.
Every one of those was available to be reused, and the pattern has recurred at much larger scale since. In September 2025 Nepal blocked major social platforms over a registration requirement — a different justification, the same instrument, an incomparably larger audience — and the blocking triggered protests of its own. The 2018 measure is where the mechanism was tested on a constituency unlikely to march for it.
India's blocking order, issued in the same season on closely parallel reasoning, went through a court rather than a ministry and produced the same outcome by a different route.
The part that still stands
Eight years on, nothing about the 2018 order suggests it changed the thing it was announced to change, and there is no measurement that would show it either way — which is itself part of the pattern, since measures introduced this quickly are rarely built with any means of evaluating them.
What remains is the precedent. A blocking power with no definition attached is not a policy about pornography; it is a general-purpose instrument that happened to be tested on pornography first, because that is the test case least likely to attract a defence.