I have lots of whistles to blow. Things where if I expose them then the report itself will be instantly attributable to me by insiders who can correlate details. That’s often worth the risks if the corporate baddy who can ID the whistle blower is in a GDPR region (they have to keep it to themselves… cannot doxx in the EU, Brazil, or California, IIUC).
But risk heightens when many such reports are attributable under the same handle. Defensive corps can learn more about their adversary (me) through reports against other shitty corps due to the aggregation under one handle.
So each report should really be under a unique one-time-use handle (or no handle at all). Lemmy nodes have made it increasingly painful to create burner accounts (CAPTCHA, interviews, fussy email domain criteria, waiting for approval followed by denial). It’s understandable that unpaid charitable admins need to resist abusers.
Couldn’t this be solved by allowing anonymous posts? The anonymous post would be untrusted and hidden from normal view. Something like Spamassassin could score it. If the score is favorable enough it could go to a moderation queue where a registered account (not just mods) could vote it up or down if the voting account has a certain reputation level, so that an anonymous msg could then possibly reach a stage of general publication.
It could even be someone up voting their own msg. E.g. if soloActivist is has established a history of civil conduct and thus has a reputation fit for voting, soloActivist could rightfully vote on their own anonymous posts that were submitted when logged-out. The (pseudo)anonymous posts would only be attributable to soloActivist by the admin (I think).
A spammer blasting their firehose of sewage could be mitigated by a tar pit – one msg at a time policy, so you cannot submit an anonymous msg until SA finishes scoring the previous msg. SA could be artificially slowed down as volume increases.
As it stands, I just don’t report a lot of things because it’s not worth the effort that the current design imposes.
Self hosting would mean I could control account creation and make many burner accounts. But there are issues with that:
I think it complicates the problem and then each author has to deal with the same. If it’s solved at the fedi API level, then the existing infrastructure is ready to work.
(edit) I recall hearing about a fedi client application that operates in a serverless way. I don’t recall the name of it and know little about how it works, but it is claimed to not depend on account creation on a server and it somehow has some immunity to federation politics. Maybe that thing could work but I would have to find it again. It’s never talked about and I wonder why that is… maybe it does not work as advertised.
I’m not suggesting you host a normal instance, I’m suggesting making a fork that makes the instance logless and allows anonymous burners.
The important thing is to be a black hole for investigating origins of posts, right?
A logless instance that does not keep track of accesses/ips/etc would mean that even in the case of a subpoena there’s nothing to turn over, and the ability to make burners with just an id and a password would ensure nothing is trackable.
You will need to make some adjustments to prevent botting but other than that this takes care of cross referencing, in my book.
Basically the same stuff that companies like mullvad do for their VPN hosting.
In this case, wouldn’t it be important for the informant and platform to be controlled by different people? It’d look mighty suspicious if you were ever suspected of whistleblowing or leaking and you just happened to run your own instance that specifically caters to that need.
Probably, but there’s clearly not a lot of interest at present, and anyone running one such instance would absolutely incur scrutiny nonetheless.
It’d probably be simpler to send the information to a trusted moderator on some specialized community through some dead drop, rather than going through the hassle of making a whole system for throwaway accounts, at that.