About
An independent collective studying what gets removed, and who asks
Torira Research is a small, independent, non-commercial research collective based in Indonesia. We were formed to answer a question that is asked often and answered rarely with evidence: when material concerning Indonesia disappears from online platforms at the request of the state, who made the request, on what legal basis, and what happened next.
Why this exists
Indonesia has an active legal apparatus for acting on online speech, and platforms operating here receive demands under it. Much of the public conversation about that apparatus proceeds on anecdote: a case that reached the news, a screenshot, a rumour of a takedown. Anecdote is not nothing, but it cannot answer questions of scale, pattern, or change over time. It cannot tell you whether a particular provision is used once a year or once a week, whether requests concentrate in a handful of agencies, or whether platforms treat Indonesian demands differently from others.
Records of these demands do exist in public archives, submitted by the platforms that receive them. They are incomplete and partly redacted, but they are real primary evidence, and they are underused in Indonesian public debate. Reading them systematically is unglamorous work that mostly consists of careful counting and honest caveats. We think it is worth doing, and we think the results belong in public.
What we are, precisely
- Legal status
- None. Torira Research is an unincorporated collective of individuals, not a registered foundation, association, or company. We describe ourselves as a collective because that is what we are, and we would rather be accurate than impressive.
- Commercial status
- Non-commercial. We sell no products or services, we take no client work, and we do not monitor content on behalf of any company, agency, or individual.
- Funding
- Self-funded. The collective receives no grants, no institutional support, and no money from any party with an interest in our findings. Its members meet the costs of the work themselves. If that ever changes, we will name the funder here.
- Working language
- We publish in English so that our findings are legible to the international research community, and we quote Indonesian-language source material in the original with translation.
- Location
- Bandung, Indonesia.
- Began work
- 2025
What we do not do
This list matters as much as the research programme, because the archives we rely on are maintained for researchers and are misused by people looking for something else entirely.
No takedown or reputation services
We do not file removal requests, we do not help anyone file them, and we do not help anyone contest them. We are not a legal service and we give no legal advice.
No brand or content monitoring
We do not track notices concerning a particular company, client, or individual on their behalf, and we decline requests to do so. Our unit of analysis is the pattern, not the person.
No republication of personal data
We do not reproduce the personal information of senders, targets, or third parties, and we do not attempt to reverse redactions applied by the archives we use.
No advocacy positions for hire
We are not commissioned by any party to reach a conclusion. Where our findings support or undermine a policy position, we report them either way.
How we expect to be judged
By whether our published methods let a stranger reproduce our numbers. By whether we state our limitations before a critic has to point them out. By whether we correct ourselves promptly and visibly when we get something wrong. Our publications page lists forthcoming work with dates precisely so that this is measurable rather than rhetorical.