I don’t really understand the legal theory here. Party A (Meta) writes a contract with Party B (a business account holder) ostensibly granting A the right to scrape B’s website and a license to the content.
But why is there any presumption that B owns the content or has any right to grant this license? It’s extremely common for a publisher to license content from a third party under terms that may not allow this sort of sublicencing. What happens if the publisher can’t grant this license?
In this case, Party B is sandwiched between Party A (scraper such as Meta) and Party C (the third-party content provider). The scraping agreement dictates A-B relationship and the licensing agreement dictates B-C relationship. Devil is in the fine print as far as what is allowed in both agreements.
Party B has already committed to uphold B-C. Entering A-B may put them at risk of violating that agreement by appearing to grant Party A rights that Party C never granted to anyone but Party B (and maybe even rights that Party B doesn't enjoy). Party B's options are don't consent to A-B (e.g., withdraw from Meta program), take down the Party C content (which would suck if they've already paid for it), or negotiate a carve-out with one or both of Parties A and C.
IANAL, but copyright binds everyone regardless of entering into any agreement. In that case Meta would still be potentially liable for its infringement, exactly because "thin-air right" is not a thing. B could not grant it, because they don't have it.
The problem there is that both B and C are pulley much smaller than Meta. Copyright and other IP law at the scale of meta has become something that only really works for those with significant resources.
AFAIK scraping itself doesn't go against copyright, only redistribution matters with respect to copyright (which is why B and C must have an agreement together).
I am not an expert in licensing legalese, but that strikes me as awfully wrong. If I am e.g. a musician and I license my work to be used by B for specific purposes and B unilaterally signs an agreement with A that gives them rights to my stuff, the question is gonna be what MY licensing terms allow. They have an agreement with me. Whether they enter other agreements is their decision to make, but they still have to fulfill my terms.
Sure, they should not sign things that put them into the spot of having to break my licensing agreement with me, but the fact that these kind of legal stretches are even seen as valid in the US reassures me in my stance to avoid business relations with the US. I prefer countries with a legal system that can't easily be bent out of shape using money.
If they really want to be spiteful they can rewrite all the news articles they post to Meta using Muse. Then we can finally see if the theories of statistical model collapse hold up at the trillion param level model scale.
I'd really be curious to see how much of Muse and similar 2026 models are actually from scraping, especially from this hypothetical unsorted random sites.
My understanding is that most of the gains after the early GPT-3 "add more weights" era have come from post-training from synthetic data, especially through reinforcement learning.
I couldn't quickly find any public statements about the actual breakdown of training data, other than almost every lab saying that post-training size was "scaling" or growing "by an order of magnitude" from their previous major release.
I'd guess that Meta is going to scrap everything anyway, and this license addition is just something that they think may come in useful if there are future court cases. The details of who has permission to grant licenses won't matter until then.
Not a lawyer, but my understanding is that crawling is a gray area.
By putting that into their terms of service that most of the planet is agreeing to, they protect themselves in the case someone complains.
Also, this is about crawling and has little to do with the license of the contents.
Not a lawyer, but my understanding is that if a Big Tech company doesn't explicitly rule out using your data, they're using your data. And if they rule it out explicitly, they will eventually still use your data.
- Find examples of content that will constitute infringement. Then send Meta legal a notice saying here are examples where the publisher does not own the content, and their scraping is not legal. Our current/future clients will be damaged by Meta's planned actions.
Is that different from them doing whatever with your data, without knowing if that's actually messages written by you, pictures of you alone taken by you etc?
Dead wrong on both points. Having a license to publish something on a website does not mean that the publisher has the right to grant sublicenses or transfer the license to someone else. For example, stock image licenses are normally non-sublicensable and non-transferable.
But Meta is not asking for anything of the sort, like you must assign to us the rights to use stock images. Just consent to the scraping activity.
The publisher using stock images does have the right to transmit those images to visitors. Those visitors don't have the right to incorporate the images into their own public work that they rebroadcast. But Meta isn't asking for that; they want consent from the business to scrape. The subsequent copyright violation is then a matter between Meta and the vendor of the stock images.
You’re missing half the picture here. It’s not the scraping that’s the big problem, it’s what the rest of the TOS say they can do with everything they get from scraping your site. Like use it to train their AI and provide responses to people questions. That’s part is the real legally contentious part, for a multitude of reasons. I imagine they see this as a way of getting around things like Australia’s requirement they pay a fee to news outlets if they show their content.
The TOS also says they can create fake accounts on your site, for scraping. Which means if you have anything on your site, that you don’t have full legal rights to distribute and use in any fashion you like, you are at significant legal risk if you continue to use anything meta after 30 October.
I'm not missing anything. I just don't believe that by granting those permissions you're making the legal representation that you are licensing third party material. The request is not worded in those terms, even, just as a consent to engage in scraping and subsequent uses of the material. For those materials that you do own, the consent does amount to licensing. For materials you don't own, the consent only covers the activity of scraping; you consenting to the material being used for training simply means nothing. You have attested that you have no objection to that. That doesn't remove the copyright holder's right to have an objection. All those things they are asking for, any visitor can engage in without any permission, and everyone knows it.
Imagine someone asks you: see that guy across the street? Do I have your consent to search his pockets? If you say yes, what does it mean? Precisely nothing.
Suppose you are asked: do I have your consent to search the pockets of everyone in this neighborhood? If you say yes, it still means nothing with regard to the guy across the street, who has not given any constent, but you've consented to your pockets being searched.
>Imagine someone asks you: see that guy across the street? Do I have your consent to search his pockets? If you say yes, what does it mean? Precisely nothing.
I think if you have some kind of authority over the place where this happens and/or if you facilitate an illegal action then you could be breaking the law.
> Because they placed the content on a public website.
I don't think this is a very good argument. For instance,
if I have a company, called 'tickle-cats', I may expect
regular customers to go and tickle a cat for, say, 1 cent,
as cost. So I would expect regular people to visit my
homepage.
I may not necessarily expect or want to have bots, scrapers
and what not, per se. I may have this by extension - possibly,
e. g. to rank a website more highly. But bots and so forth
are not my real target audience to begin with (and some are just malicious, trying to disrupt the business); some company owners may want to not have to do anything with them (bots, crawlers etc...). That's just one example of many more;
I don't think the distinction "public versus private" is the only
one that can be applied here.
This is some next level legal judo. Say what you want about the metaverse or muse, but it's clear that Meta is still innovating when it comes to its Terms of Service.
No, it's the pick-up artists. They're the ones that know how to find ways to get around saying "no" to touch what they wanna touch even if the rest of us didn't want them to.
Those are certainly among the most interesting engineers. For example, what were the American Founding Fathers, if not "legal engineers" (and certainly many of them were also "financial engineers", given this was one of their most pressing needs).
I guess if you find yourself preferring the status quo all the time they might be scary but given I can't think of any time in history where the status quo was totally desirable I'm not sure what there is to be scared of exactly.
My interpretation was that they’re referring to legal/financial engineers in the context of corporate butheshareholderism, where rather than trying to e.g. build a good nation, the goal is to maximize profits.
These are two types of engineers that, at some places, aim to find loopholes that immensely help themselves while immensely hurting others, maybe as collateral, maybe by design. The fact that you’re so gung-ho about this…is alarming.
Doesn't them requiring a license from partners to train on their content undermine the more general argument that they don't need a license from copyright owners to train on copyrighted content -- which I assume most big AI firms would love to make?
IANAL, but this looks like Meta lawyers trying to shield themselves against partners' suing them, but enough cases like this could help set a precedent that -- yes, you do in fact need a license.
If you don't have a license to download the material, aren't you doing piracy?
If the licenses say that download is only permitted if you agree not to train on it or similar, then if you have downloaded it, you have agreed to the license, have you not?
TOS has nothing to do with copyright. A TOS can prevent scraping, training, or any other activity that is not covered by copyright. Public web scraping is subject only to copyright because no TOS has been agreed to, but what Meta is doing here applies to scraping things that are not available on the public web.
What happens when I put a Terms of Service on my website (esp. with a firm click-wrap) which says Meta agrees to not scrape my website--notwithstanding any prior agreement--and will owe me a billion dollars in damages if they do?
(But my guess is that such Terms of Services are not considered accepted only by visiting or crawling the website. There probably need to be at least a clock through performed by an human)
How about if you put that notice on the website, but also write it out explicitly and post it via registered mail to the Meta lawyers. Explicitly state that the content on your site is not licenced to be scraped for free and any usage must be paid for (eg $10/GET with penalties for non-payment), and that any usage of the site by Meta or their agents is an indication that they accept the terms fully.
Then check your logs and start sending appropriate invoices.
That's a bold move. I imagine a lot of businesses will leave Meta over this. And if their gamble pays off, and businesses can't afford not to advertise on Meta platforms, then there is probably a good antitrust case for them abusing their oligopoly power.
So basically, thanks for building our business side of Facebook for all those years, we are changing the rules of the game now in so if you want to continue using our service, we get to use all of your website's data however we like in perpetuity.
There isn't even a veil of "customer satisfaction is important" anymore.
No, but you do in order to rebroadcast some content they created themselves. If they obtain this consent from a business, then that business cannot find some piece of their copyrighted stuff in Meta's AI output and launch a copyright complaint.
Do business accounts on FB provide user activity data, such as checkout completion, user acquisition funnels, or other useful labeling of web behavior?
Because the scrapes become more interesting once combined with such information.
I assume this is because of Muse? A lot of businesses block bots which stops Muse from being able to buy from those websites on a user's behalf. It seems like that'd be good for the businesses too.
This makes the most sense. Meta can guarantee at least that cohort of sites will permit their agents to operate, and all they needed to do was publish a tos update. Some Meta PM is smiling today, in between bites of infant flesh.
iirc when I was dealing with reviewing traffic and AI scraping bots recently, I noticed something about Meta's: When a user links to a page, the bots that grab oembed data to generate the link preview are using the same User Agent string as the Meta AI crawlers that scrape content.
For context, oembed data about your page could be seen as information you're basically happy to share, so other sites can create things like previews.
That's not necessary, there are other companies that just use different UA strings for the different tasks. They appear to deliberately make it so that if you have nothing to do with Meta and block their content scrapers, you also break basic functionality for people who happen to share a link to your site in their products.
And that they let businesses change the license at any time, with just an email notification, and the only way to opt out of the new terms is to stop using the service.
How else would you "opt out"? You want to have your cake and eat it too?
Most people don't care about the Terms of Service. Would you have a more positive assessment of the judicial system if it infantilized its citizens? If someone doesn't read it the first time or ignores an email what is the system supposed to do? Stand over your shoulder and make you read it? Do you want citizens to be prosecuted for signing things on the theory they didn't actually read them?
What number of people actually read the terms, actually read the new terms, decide they don't like the new terms, but after caring this much decide they don't want to "opt out" (I'm still not clear what remedy you want to invent for this)? Do you think this number is so high that you should be shocked that for some reason this is not a priority? I'm not even sure you've actually articulated a problem.
Is there a Terms of Service change you find particularly offensive that might better explain your concerns here? If so, I would be interested to hear you describe it with greater specificity.
> How else would you "opt out"? You want to have your cake and eat it too
If the TOS changes you should be automatically opted out until you explicitly accept the new agreement
That way if you stop using the service because of the new TOS, it actually means your old activity is still bound by the previous TOS not the new one. Weird how that's never an option though.
> If the TOS changes you should be automatically opted out until you explicitly accept the new agreement
Does it not work this way? Don't the terms you sign often say something like these terms can be updated at any time? Yet they still often want you to agree to the new terms to proceed, right?
> That way if you stop using the service because of the new TOS, it actually means your old activity is still bound by the previous TOS not the new one.
Are you suggesting a statutory remedy? What is the average diff of a terms change? Are they really that meaningful? I would think the scope of what they can actually change is controlled by various other areas of law. If you don't accept the new terms are you concerned your old data is going to be used in a new way, while you remain opted out of the future? Is that actually how it works? It feels like it would not, which I gather is why they often want you to explicitly accept the new terms.
Once again it would be helpful if a specific change or event was brought up here that better explained your concerns. I feel like there are times when I am forced to agree to new terms to continue using the service, and other times where it feels like I am not. I would suspect this relates to some controlling law regarding the nature of the update itself, as not all updates are equally meaningful?
I don't know about the USA, but it seems that generally filming in public where you catch sight of passerbys and filming a specific person in public spaces are wildly different things.
If I stand outside your house and follow you all day recording your every moment in a public space, is that legal or has that turned into harassment? Certainly in many countries, filming an incidental person may be legal but filming them in detail and from every angle is not (eg upskirting)
This is an agreement between two businesses, not a clickwrap license being forced upon consumers. Businesses have plenty of alternatives to Meta; they can just close their accounts if they don't like the terms.
This company is poison. You have built a few things people liked, bought some other things people really liked, and then used that good will to completely screw over the entire internet.
I know why you haven’t quit yet, but you probably should right? You’ve got your vested shares and you’re comfortable. Why are you still doing this to us?
One could argue that if you hate these megacorps then the best thing you can do (as an engineer)is to work for them and purposefully be a monkey wrench in the cogs. Be competent enough to not get fired, maximize the company resources you take up while minimizing your output. Delay delay delay and if you're really good at it, never deliver. Take as much of their time and money as possible and give as little back. Thats basically what they're doing to the rest of us. And when its gets too evil you can always leave. In the meantime you are taking up headcount that could have actually advanced their goals further, faster, evil-er.
As a business owner who has meta accounts and even advertises on meta, I have long since blocked their crawler traffic. They try to hide their scraper traffic dishonestly behind user agents that purport to be Facebook users browsing the web via their Facebook browser, the thing is, I know it’s a lie because there is absolutely no reason for 1000s of Facebook users from Ireland to want to know anything about a small local business in Canada, that only services local clients within 50km radius due to the nature of the business. So I know it’s a lie, therefore was blocked. And guess what, fuck your terms of service the blocks are staying in place.
IANAL and I know I'm a minority opinion, but: seems to me that scraping must be legal. If you put content online, it is readable by anyone (or anything) who cares to look. Don't want it read, put it behind a login or a paywall.
But why is there any presumption that B owns the content or has any right to grant this license? It’s extremely common for a publisher to license content from a third party under terms that may not allow this sort of sublicencing. What happens if the publisher can’t grant this license?
Party B has already committed to uphold B-C. Entering A-B may put them at risk of violating that agreement by appearing to grant Party A rights that Party C never granted to anyone but Party B (and maybe even rights that Party B doesn't enjoy). Party B's options are don't consent to A-B (e.g., withdraw from Meta program), take down the Party C content (which would suck if they've already paid for it), or negotiate a carve-out with one or both of Parties A and C.
(IANAL)
Sure, they should not sign things that put them into the spot of having to break my licensing agreement with me, but the fact that these kind of legal stretches are even seen as valid in the US reassures me in my stance to avoid business relations with the US. I prefer countries with a legal system that can't easily be bent out of shape using money.
My understanding is that most of the gains after the early GPT-3 "add more weights" era have come from post-training from synthetic data, especially through reinforcement learning.
I couldn't quickly find any public statements about the actual breakdown of training data, other than almost every lab saying that post-training size was "scaling" or growing "by an order of magnitude" from their previous major release.
Meta's past behaviour has been to ignore illegality in business, with the hope that they can either win in court or stave off legal action altogether.
That's how they acted with the books they pirated to train models and when they targeted children with gambling products to increase revenue.
They love a legal grey area. It's not what's moral it's what you can get away with.
Also, this is about crawling and has little to do with the license of the contents.
What happens if the publisher has a license granted by their tier1 but their tier1 can’t grant this license?
- Find examples of content that will constitute infringement. Then send Meta legal a notice saying here are examples where the publisher does not own the content, and their scraping is not legal. Our current/future clients will be damaged by Meta's planned actions.
This might elevate it to willful infringement.
Because they placed the content on a public website.
> What happens if the publisher can’t grant this license?
Then it's unlikely that the content is something they themselves have the rights to broadcast on their website.
The publisher using stock images does have the right to transmit those images to visitors. Those visitors don't have the right to incorporate the images into their own public work that they rebroadcast. But Meta isn't asking for that; they want consent from the business to scrape. The subsequent copyright violation is then a matter between Meta and the vendor of the stock images.
Imagine someone asks you: see that guy across the street? Do I have your consent to search his pockets? If you say yes, what does it mean? Precisely nothing.
Suppose you are asked: do I have your consent to search the pockets of everyone in this neighborhood? If you say yes, it still means nothing with regard to the guy across the street, who has not given any constent, but you've consented to your pockets being searched.
I think if you have some kind of authority over the place where this happens and/or if you facilitate an illegal action then you could be breaking the law.
I don't think this is a very good argument. For instance, if I have a company, called 'tickle-cats', I may expect regular customers to go and tickle a cat for, say, 1 cent, as cost. So I would expect regular people to visit my homepage.
I may not necessarily expect or want to have bots, scrapers and what not, per se. I may have this by extension - possibly, e. g. to rank a website more highly. But bots and so forth are not my real target audience to begin with (and some are just malicious, trying to disrupt the business); some company owners may want to not have to do anything with them (bots, crawlers etc...). That's just one example of many more; I don't think the distinction "public versus private" is the only one that can be applied here.
Those are certainly among the most interesting engineers. For example, what were the American Founding Fathers, if not "legal engineers" (and certainly many of them were also "financial engineers", given this was one of their most pressing needs).
I guess if you find yourself preferring the status quo all the time they might be scary but given I can't think of any time in history where the status quo was totally desirable I'm not sure what there is to be scared of exactly.
Could you be more specific?
They're shifty fuckers enabling greedy scummy fuckers to make everything shittier forever for everyone in the world except themselves.
IANAL, but this looks like Meta lawyers trying to shield themselves against partners' suing them, but enough cases like this could help set a precedent that -- yes, you do in fact need a license.
If the licenses say that download is only permitted if you agree not to train on it or similar, then if you have downloaded it, you have agreed to the license, have you not?
(But my guess is that such Terms of Services are not considered accepted only by visiting or crawling the website. There probably need to be at least a clock through performed by an human)
Then check your logs and start sending appropriate invoices.
Then after 1yr of unpaid invoices: <profit>
Well, assuming, y'know, justice under the law and such.
There isn't even a veil of "customer satisfaction is important" anymore.
Because the scrapes become more interesting once combined with such information.
https://www.cbc.ca/news/politics/online-news-act-meta-facebo...
For context, oembed data about your page could be seen as information you're basically happy to share, so other sites can create things like previews.
That's not necessary, there are other companies that just use different UA strings for the different tasks. They appear to deliberately make it so that if you have nothing to do with Meta and block their content scrapers, you also break basic functionality for people who happen to share a link to your site in their products.
> "[T]hey'll give people some freedom, but essentially it will be 20th-century capitalism: a corporate elite protected by laws and tax-codes."
Most people don't care about the Terms of Service. Would you have a more positive assessment of the judicial system if it infantilized its citizens? If someone doesn't read it the first time or ignores an email what is the system supposed to do? Stand over your shoulder and make you read it? Do you want citizens to be prosecuted for signing things on the theory they didn't actually read them?
What number of people actually read the terms, actually read the new terms, decide they don't like the new terms, but after caring this much decide they don't want to "opt out" (I'm still not clear what remedy you want to invent for this)? Do you think this number is so high that you should be shocked that for some reason this is not a priority? I'm not even sure you've actually articulated a problem.
Is there a Terms of Service change you find particularly offensive that might better explain your concerns here? If so, I would be interested to hear you describe it with greater specificity.
If the TOS changes you should be automatically opted out until you explicitly accept the new agreement
That way if you stop using the service because of the new TOS, it actually means your old activity is still bound by the previous TOS not the new one. Weird how that's never an option though.
Does it not work this way? Don't the terms you sign often say something like these terms can be updated at any time? Yet they still often want you to agree to the new terms to proceed, right?
> That way if you stop using the service because of the new TOS, it actually means your old activity is still bound by the previous TOS not the new one.
Are you suggesting a statutory remedy? What is the average diff of a terms change? Are they really that meaningful? I would think the scope of what they can actually change is controlled by various other areas of law. If you don't accept the new terms are you concerned your old data is going to be used in a new way, while you remain opted out of the future? Is that actually how it works? It feels like it would not, which I gather is why they often want you to explicitly accept the new terms.
Once again it would be helpful if a specific change or event was brought up here that better explained your concerns. I feel like there are times when I am forced to agree to new terms to continue using the service, and other times where it feels like I am not. I would suspect this relates to some controlling law regarding the nature of the update itself, as not all updates are equally meaningful?
If I stand outside your house and follow you all day recording your every moment in a public space, is that legal or has that turned into harassment? Certainly in many countries, filming an incidental person may be legal but filming them in detail and from every angle is not (eg upskirting)
I know why you haven’t quit yet, but you probably should right? You’ve got your vested shares and you’re comfortable. Why are you still doing this to us?
https://en.wikipedia.org/wiki/Bundling_(antitrust_law)
Oddly enough, many don't seem to want to have a company homepage, in particular smaller companies.