I recently learned that voting on lemmy is not anonymous. Anyone can get information about who has upvoted and downvoted a post or comment.
In combination with your IP, this is a massive privacy (maybe even physical security) risk. Also, people can target you for your votes.
Sadly, this is something where I would prefer Reddit over Lemmy. Big tech scrapes data from both places anyways, at least Reddit is safe.
I don’t think IP addresses federate? I think only your instance admin can see your IP address. In any case, though, you should generally always assume that your up/down votes on any service are recorded and tied to your username. If you can come back later and change your vote, that vote is tied to your username. It may not be visible to other users, but the server admins can absolutely see what you’re doing.
Reddit might not make your votes publicly visible, but they’re absolutely tracking them and using that information to select what you see, including advertising. They might not directly share those votes with advertisers, but they almost certainly are sharing your interests based on your votes. And you should assume Reddit and others will comply if the government comes asking for what users liked a post the government opposes, or who downvoted a post praising a new government initiative.
It depends on your threat model, but your threat model might change. Freedom of speech might be curtailed by politicians even when that’s supposed to be unconstitutional. What might be safe to do online now might become unsafe in a year or two.
YSK: every action you take online, even as simple as an Upvote or Like, might be recorded and may come back to haunt you
Reddit is one entity, and by providing a service it is bound by a variety of privacy and data protection regulations. On the fediverse anyone can accumulate any of that information and store it for years, and they are not bound by any such data management or privacy laws. It’s absolutely shocking to me that a place which is otherwise quite obsessed with privacy just brushes aside this distinction. As it stands a vote on the fediverse is far more likely to have real consequences versus one on reddit if, say, ones phone is searched at a border.
This could be mitigated considerably with simple voting agents, as piefed tried to do, but this idea was killed by idiotic forum politics over fears of “vote manipulation.”
Yes, this is not hyperbole - the otherwise “privacy focused” leaders of the fediverse are more concerned with fake Internet points than real privacy concerns.
