> Pavel Durov, the inventor of Telegram, stated on his personal channel that the material sought by the Brazilian court is “technologically impossible” to get.
> Durov also stated that Telegram has already left nations such as China, Iran, and Russia (where the app was established) owing to local rules, and that the software may leave Brazil as well.
> According to the app’s creator, leaving a nation is “preferable to the betrayal of our users and the beliefs we were founded on.”
i know the people involved in creating the report that caused the request. it's a group on federal universities (brazil has state owned universities) who monitored OPEN groups discussing politics to measure impact of known fake news on presidential election.
remember, they are mostly phds focused on publishing what they find. so everything is under ethics committee... they didn't even used data from groups that require a invite. only 100pct open groups. also they were not measuring complex fakenews metrics. only widely accepted ones. flat earth level.
now i haven't seen the request, but i doubt it's the e2e stuff, if telegram even have that...
The police asked them for lots of information like CPF, bank accounts, home addresses. Stuff Telegram never asks anyone about. I assumed that's what they meant by impossible.
Also, wait a second. You know people who are monitoring others to study the impact of disinformation? Where can I find more information about these activities? Are you monitoring this place? Am I being monitored?
you will find it when published. assuming you read anything other than facebook.
big advantage of public education system is that research is, well, public.
also yes. i assume. everything public today goes to some database. If nothing else just to train some autocomplete people will worship as god.
learn basic privacy and security ops if you care about it in a public space, instead of just being melodramatic.
lastly, i doubt they ask for cpf, unless they have a paid service that takes Real, then they should have a cpf on file otherwise it's wire fraud equivalent in brazil law. meh. i assumed they were just asking ip addresses. not that i care much about any of that.
> learn basic privacy and security ops if you care about it in a public space, instead of just being melodramatic.
What I care about is whether researchers are building lists of wrongthinkers. Lists that are obviously going to be weaponized once the censorship laws inevitably pass.
If it's a proper academic study, then surely the methodology was evaluated by an ethics committee. If your study generated the report that led to Telegram being banned, then the information being collected must not be anonymized.
The WhatsApp monitoring system paper mentions partial phone numbers are stored "for geographical analysis purposes".
> i doubt they ask for cpf
They don't. Hence the impossibility of complying with the judge's order.
The big problem is that the "monitoring" is done by opaque and questionable machine learning algorithms. The dilemma is that at the same time such algorithms are the only option because there is so much garbage out there (and with LLMs there is going to be much more).
6 comments
[ 1.3 ms ] story [ 24.3 ms ] thread> Durov also stated that Telegram has already left nations such as China, Iran, and Russia (where the app was established) owing to local rules, and that the software may leave Brazil as well.
> According to the app’s creator, leaving a nation is “preferable to the betrayal of our users and the beliefs we were founded on.”
Interesting. I thought they were going to comply.
i know the people involved in creating the report that caused the request. it's a group on federal universities (brazil has state owned universities) who monitored OPEN groups discussing politics to measure impact of known fake news on presidential election.
now i haven't seen the request, but i doubt it's the e2e stuff, if telegram even have that...Also, wait a second. You know people who are monitoring others to study the impact of disinformation? Where can I find more information about these activities? Are you monitoring this place? Am I being monitored?
big advantage of public education system is that research is, well, public.
also yes. i assume. everything public today goes to some database. If nothing else just to train some autocomplete people will worship as god.
learn basic privacy and security ops if you care about it in a public space, instead of just being melodramatic.
lastly, i doubt they ask for cpf, unless they have a paid service that takes Real, then they should have a cpf on file otherwise it's wire fraud equivalent in brazil law. meh. i assumed they were just asking ip addresses. not that i care much about any of that.
Which journal?
> assuming you read anything other than facebook
I don't use Facebook.
> big advantage of public education system is that research is, well, public.
Which universities? UFRJ, UFMG? I found news matching what you described dating back to 2018.
Actually, I think I found the system itself. Will definitely be checking it out later.
https://ufmg.br/comunicacao/assessoria-de-imprensa/release/f...
https://ufmg.br/comunicacao/noticias/projeto-eleicoes-sem-fa...
https://www.eleicoes-sem-fake.dcc.ufmg.br/
https://www.eleicoes-sem-fake.dcc.ufmg.br/assets/articles/we...
https://www.monitor-de-whatsapp.dcc.ufmg.br/
http://www.monitor-de-telegram.dcc.ufmg.br/
https://adanalystplus.imag.fr/pt/
> learn basic privacy and security ops if you care about it in a public space, instead of just being melodramatic.
What I care about is whether researchers are building lists of wrongthinkers. Lists that are obviously going to be weaponized once the censorship laws inevitably pass.
If it's a proper academic study, then surely the methodology was evaluated by an ethics committee. If your study generated the report that led to Telegram being banned, then the information being collected must not be anonymized.
The WhatsApp monitoring system paper mentions partial phone numbers are stored "for geographical analysis purposes".
> i doubt they ask for cpf
They don't. Hence the impossibility of complying with the judge's order.
https://scholar.google.com/scholar?hl=fi&as_sdt=0%2C5&q=arxi...
The big problem is that the "monitoring" is done by opaque and questionable machine learning algorithms. The dilemma is that at the same time such algorithms are the only option because there is so much garbage out there (and with LLMs there is going to be much more).