this post was submitted on 17 Dec 2023
47 points (94.3% liked)
Fediverse
17849 readers
1 users here now
A community dedicated to fediverse news and discussion.
Fediverse is a portmanteau of "federation" and "universe".
Getting started on Fediverse;
- What is the fediverse?
- Fediverse Platforms
- How to run your own community
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
they can already download what people are sharing. that's not what this is about.
if they want to train AIs on your data they can scrap it from the regular feed without any integration. If they needed federation to do it they could just set up a mastodon instance to grab the data - but they don't.
They can not legally just download and use what people share. Meta is under high legal scrutiny and this would be clearly against copyright.
If they can train AI on it is currently an open legal question, but for sure they can do it if you allow them to do it by federating openly with them.
if they run a mastodon instance, you'd be federating with them just as much.
are you sure these comments are somehow protected by copyright and they can't use them? when I post publicly like this I have no expectation of control over use.I'd be very surprised if I could somehow sue a company using this comment to train AIs.
I'd also be surprised if that status changed somehow if the server was then connected to threads?
If they run a Mastodon instance they would be defederated as well.
And yes you retain copyright over what you write automatically and Meta can't use it legally just like that. It doesn't matter how you feel about it, the only thing that matters is what is written in the ToS of your instance which you agreed to when signing up. Usually it has a clause that allows them to forward messages to other federated instances, which would include Threads unless defederated.
Training AI is a big exception to all this as it is currently not known how to deal with all this legally, as training an AI does not require to copy the content but rather just have the training algorithm "look" at it...
I can't find anything like what you're describing - my instance has a legal notice which is just a disclaimer saying they can't be held liable, lemmy.world has a fair use and terms of service which are 404s and their privacy policy just says they won't sell your data (but might use it for internal research) - can you tell me what you mean?
Well, if they don't have a ToS clause for that, then technically they are violating your copyright by sharing your contributions with other instances.
Most commercial services force users to completely sign over the legal rights for their contributions to the service.
On the Lemmy instance I am on the ToS clearly states that people agree to have their original contributions licensed under the Creative Commons Attribution Share Alike 4.0 or later license, which allows redistribution if certain terms are fulfilled.
It only takes looking at your data to figure out your trends, save the trend, and serve you ads.
Think about it: public posts are public. It's the same as you putting a note in the town square. Anyone can look at it and see the username of who wrote it.
Defederation doesn't stop that, it just inconveniences people who want to use/see both sides from one login.
This is not how it works technically.
For Meta to analyze your data they need to either scrape it (legally questionable and scraper bots are commonly blocked on server level) or work with a local copy. By federating with them you are allowing them to legally make a local copy of all the posts of the instance.
Newspaper articled are often also public, yet google got sued (and lost) because they were scraping and analyzing them to put previews in their search results.
Just because something is public doesn't mean you can just take it. Copyright still aplies.
Defederation does stop legal use, and Meta is already in enough legal trouble, especially in the EU, that they are unlikely to blatantly pirate user contributed content from sites that defederated from them.
I highly doubt it. The laws haven't caught up to what you're saying. Basically what you are saying would make scraping illegal.
As far as I know it isn't. If it is: please cite a published law article or something similar discussing it.
You are arguing against something I never claimed.
I said that if Facebook wants to copy and republish something (so that they can put advertisements next to it) they will do that through legal ways as posts are copyright protected. The only way they can do that is through openly federating instances that allow republishing in their ToS.
It is totally irrelevant if scraping is legal or not (its a gray area), the questing is rather does defederation stop Facebook from using posts from the Fediverse, and it likely does (IANAL).
We're seemingly talking about different things. I don't think they can put ads next to my content if they scraped it... Then again isn't this how Google works? They even have caches of a lot of the content so you don't need to hit the original one... So we know they store the pages.
I see how if federated it's more of a gray area since it's federated: so maybe they can put ads? Idk seems like another gray area. I wonder how a ToS can be applied from a legal perspective if the content was federated instead of directly posted. Then again Google just looks at a robots.txt file to figure out what/how to scrape. Maybe that should apply here somehow? Idk.
I'm guessing it'll take many years for laws to catch up... And they'll be written by whoever has more money at the time.
Can't they? The internet is full of scrapers and reposters, and I don't think I've ever seen a company like that going down. I'm not sure what the point would be (generate a dataset? perform sentiment analysis on certain topics? streamline their tag detection system with off-platform data?), but they can, as long as they follow the relevant privacy laws ("practically no restrictions" outside of the EEA+UK+California, "anonymise before processing" everywhere else).
They would only violate copyright if they redistribute the content. Downloading and processing the content offline wouldn't really be breaking any laws, outside of a structured PII situation.
You are missing the point entirely: in order to put advertisement next to it, they of course need to copy and redistribute it.
Why do you think they haven't enabled two way federation yet? It is precisely because of the unclear legal situation regarding content sourced through federation.
And as shady as Meta is, they are an established company with a big legal department and not some web scalper operating from a 3rd world country.