Europe's idea of conservative is what Americans would call far left. It's all relative anyway. If you're just trying to conserve something left wing against something even farther left wing then it's still left. Mostly meaningless label. Almost no one even knows where the political terms left and right came from. People call Afghanistan right wing as if that means something.
> "Of the deleted comments examined across platforms and countries, between 87.5% and 99.7%, depending on the sample, were legally permissible. The highest proportion of legally permissible deleted comments was observed in Germany, where 99.7% and 98.9% of deleted comments were found to be legal on Facebook and YouTube, respectively.
Wow, that's quite a bit worse than I would have guessed even as a critic of these types of laws.
People see the instances of megacorps blatantly flouting laws and assume this must be their default behavior, but really the default is typically conservatism and caution with regards to regulations (with an occasional flout where they feel they can get away with it or they can make a ton of money off of the flouting).
This is an area where it's gonna be pretty subjective whether many comments are law breaking, so the internal guidelines for censors will likely err on the side of caution, which means removal.
>People see the instances of megacorps blatantly flouting laws and assume this must be their default behavior, but really the default is typically conservatism and caution with regards to regulations (with an occasional flout where they feel they can get away with it or they can make a ton of money off of the flouting).
This is exactly the worst hypocrisy of these very same megacorps, and particularly notable in mega tech corps, which have a particular tendency towards preaching loudly while practicing very differently.
On the one hand, they play themselves as arbiters of moral values and display a nearly (no scratch that, overtly, blatantly) preachy attitude towards their users and customers and even the wider public about social responsibility, fighting "incorrect" thinking or conduct, following proper regulations and etc and etc. With this attitude in mind, they essentially forcibly (at least within their reach as platforms) "guide" average people into certain types of behavior. They also make public speeches, blog posts, PR releases and all sorts of endless vomit about the de jure moral or social crusades of our time, depending on each company's specific context.
Aaand then, they also simultaneously perform nearly every kind of legal, social and financial malfeasance you can imagine, often blatantly for the sake of their real bottom line, which is market dominance and more financial power. (looking at just a few examples: OpenAI's blatantly predatory behavior toward pretty much everyone else's content, or its obviously dishonest crap about the supposed dangers of AI in its efforts to scare regulators in its favor, or Meta's too-numerous-to-count scandals across many years, or Google's intense quest to utterly destroy search for the sake of further ad dollars, or Amazon's indifference to an absolute avalanche of consumer fraud on its product sales platform while at the same time squeezing sellers as often and much as possible)
All of the above is to not even mention the completely presumptuous attitude among nearly all modern tech companies, and more recently, most other major consumer-oriented companies of all kinds (these days any large company is really a tech company at heart) that they have an absolute right to not only pry into everything you do no matter how private, but also store and distribute or sell that information as they please to nearly anyone at all with zero consequence. This last has been taken to the point where for all their talk of "valuing your privacy" the default behavior is the diametric opposite, and the default response to anyone who really tries to dial it back is overt hostility.
It's all quite grotesque.
How anyone could take seriously a single major major techcorp leader's preaching on any social issue is beyond me.
And indeed, is the think tank going to fund the lawyers to argue the point if Meta leaves a post-up and the nation of Germany brings suit? Or is the think tank going to find Meta new users if Meta leaves the post up and individuals decide to stop using Facebook because they don't want to see that crap?
Companies drop posts for violations of community terms that have nothing to do with what is legal.
According to the article, most removed comments were about current events. You can access the report here[1][2] and see if you think the comments they highlight comport with European speech laws or not.
If you check their methodology, their coding guidelines were developed by legal experts who are native speakers in each of the target countries, so to answer "Says who?", lawyers.
The article does not state that the removed comments were removed for legal reasons.
The ad industry tends to be a lot more sensitive to certain content than governments are. (For example simple insults can easily get a Youtube video demonitized.)
There's a good chance that the removal of comments regarding killing and war is caused purely by ad regulations.
Without further information, these numbers tell nothing about the actual effects of censorship laws.
Please note that CSAM is categorized as a "legally permissible expression of general opinion" as defined by this study. I'm not sure we can draw too many conclusions about the reason for deletion here.
> Comments fell into this category if meaning could be interpreted from them, they were not illegal, did not contain linguistic attacks or fell outside the scope of the report (e.g. CSAM, revenge porn, marketing). Most comments belonged to this category.
They are saying that those CSAM fell outside of the scope of the report not that it was considered legal speech. Its awkwardly worded.
If you read further they say:
> Criminal restrictions on content that can be classed as The Future of Free Speech “revenge porn” or Child Sexual Abuse Material (CSAM) are discounted from discussion in this report as they are universally recognized as unprotected speech.
No, they are saying that since CSAM fell outside of the scope of the report that for the purposes of this report it is not coded as illegal speech.
Section 2.3 Coding of Data:
> These legal notes are based on legislation related to freedom of expression, which means that violations of marketing legislation, for example, were not coded as illegal.
If I were to begin this response with a string of insults and swear words that would turn your ears red questioning everything from your continence to your sexual history nothing about it would be illegal but it would be unfit for this forum and it would be fitting to remove it.
If I were to launch into a tirade about a particular race (pick one) if I phrased it reasonably enough to avoid falling afoul of hate speech laws it would be legal. It would also still be hate.
If I were to spread heinous lies about a group to inspire stochastic terrorism most places it would be legal.
Legality is a meaningless standard to apply to this situation.
Most content moderated by most sites isn't strictly illegal. On Hacker News you can be modded down for being impolite, or insufficiently substantive, for mere disagreement, for being repetitive or any number of reasons including "just because," since there are no real guidelines for voting here, and people tend to use flagging as a "super downvote" rather than for content that just violates the ever increasing list of guidelines.
The number of sites that limit moderation to strictly illegal material are few and far between, and they tend to degenerate into cesspools where only the worst elements prefer to gather. Even Gab recently learned this lesson[0,1].
They don't know why they were deleted. They don't know who they were deleted by. They don't actually know how many automatically-deleted bad comments they missed because they rely on scraping every ~10 minutes. It's flaw on flaw on flaw.
Sure but these platforms are not very examinable. Does that mean they should get a pass and escape accountability? I think it’s okay to assume the worst in the face of secrecy.
No, but when you don't even know if it's a platform or a volunteer community moderator, or even the user deleting their comment, making any claims is flawed.
You can distil the factual elements of the report to: "Comments that were available to us are now no longer available to us, but we don't know why"
> Does that mean they should get a pass and escape accountability?
What? My pointing out extreme flaws in data collection and analysis shouldn't mean anything other than there are problems with this report. Why are you trying to extrapolate a motive for me?
As a EU citizen I think it would be neat if I had any sort of tangible representation or means of affecting the policies and decisions made by the EU or really if there was just an debate about these things where people could be heard :-/
It's an election for the parliament, which holds a fraction of the power in the EU. The executive branch of the EU has basically zero accountability to its citizens.
Democracy isn't transitive though. If a French politician does something to cause issues in Greece, there's nothing the greeks can do about it, and no reason for the french voters to care, since their national vote is primarily about local politics.
When was the last time the implications for the European Council was even mentioned in the context of (any) local election?
The problem is that there is no way to approve of a government locally, but disapprove of their performance in the EC; and given the two conflicting concerns, virtually everyone will favor the former over the latter. This means that de facto there is no voter accountability for the executive branch.
There's also no way to approve of a politician's position on A, while disapproving of their position on B. Also known as "holding your nose" while voting. By your measure, nothing short of a direct referendum on a line-by-line basis would be democratic.
If you accept the premise of illegal speech or that certain expression doesn't meet the bar for protected speech in the first place — as the article's author does and as I do — then you've already crossed the rubicon. The how you enforce it is an implementation detail.
Deputizing big tech to handle copyright disputes is in relative terms completely uncontroversial and is exactly the same means of censorship. Regulating the world's largest megaphones is the natural choice. It's hard to argue against any of the pieces when most people accept them individually but get uncomfortable at the reality of what it means put together.
Indeed. The only quibble I might
entertain is that other mechanisms of law enforcement involve an open in public process of trial and some kind of check against absolute authority. Generally, such checks do not exist within the fiefdoms of individual social media outfits and their terms of service.
I recognize it as a wart but I'm really not sure how to address it in an international context where there is no one frame of law that can be applied, nor should there be.
""Legal online speech made up most of the removed content from posts on Facebook and YouTube in France, Germany, and Sweden,""
This whole article is idiotic. It can be re-phased as: "Meta and Alphabet don't want to hire large compliance teams, so they simply accept every request as valid." The companies have the option to push back, but they prefer not to, because they don't want to be bothered. And if a take-down request is not challenged, then we don't actually know if the courts would have found the speech illegal or not. If the courts frequently upheld the speech, then the regulations would naturally shrink in scope. But if the courts instead agreed that the speech violated the law, then the speech actually was illegal. But establishing the exact boundary between legal and illegal is something that can only be done if the companies, or some free-speech CSA, are willing to take some of these cases to court.
The criticism here is similar to the one made against content hosting companies who take down any content flagged by a DMCA request, even when the entity making the DMCA request is not the copyright holder. When the content hosting companies demonstrate cowardice regarding DMCA, then the DMCA can become a tool of abuse, but some of the criticism regarding the situation around the DMCA should be aimed at the cowardice of the companies.
It's worth repeating, too, Meta and Alphabet always have the option to leave the EU market if they find the regulations in the EU market to be overly burdensome. But if they stay in the EU market, then they'll have to abide by EU regulations.
This appears to be a report based on looking at all posts that have been removed for any reason and complaining that of the vast majority, the law did not compel their removal - and therefore the law is too harsh?
There's a bit of handwaving about how the law is chilling free speech by encouraging companies to remove it too aggressively, but I don't think it's very convincing.
If I’m not mistaken, companies that facilitate or host discussion are given the legal burden of policing speech online but there is obviously no consequence for removing more than necessary. It’s very rational for them to err on the side of caution and remove anything that has the slightest chance of getting them in trouble and being forced to pay a fine. Conversely, bad acting companies can remove content they disagree with under the guise of compliance.
This is just another example of poorly targeted legislation with negative externalities that will far eclipse whatever problem is being solved.
The bigger problem is that we as a society for some godforsaken reason believe it’s best to change social behavior with law. This is the same shit we have been doing for centuries and it has never worked. You cannot stop people from internalizing being a human turd, although social pressures can. But if all you have is a hammer…
I support the message about allowing legal speech but this is ridiculous:
> Three actors have the ability to delete a comment: the user who posted it, the social media platform, or the owner of the channel or page, where the comment was made. This report does not have the capability to distinguish between these three scenarios. [..] Therefore, this report treats the three reasons for comment disappearance as a single phenomenon.
There's no connection at all with EU laws if they include the posts I delete myself.
They chose media and political pages, found 3% of comments were removed (without determining the reason for removal), and then had lawyers review the deleted comments, finding most of them fell into the catch-all category defined as follows:
> Comments fell into this category if meaning could be interpreted from them, they were not illegal, did not contain linguistic attacks or fell outside the scope of the report (e.g. CSAM, revenge porn, marketing).
If I'm reading this right, this contains content such as "Buy Coke!"
Is this useful data?
I certainly don't think it supports their stated position:
> In a world that works the way policymakers intend, we would expect to find that most deleted comments constitute illegal speech.
I'm not sure I agree with this position in the abstract either.
I've seen skyscrapers with fewer flaws than the report this is based on.
But seriously, the data is scraped. In my experience, these networks are pretty finicky about which comments they'll actually show you (even if they're not deleted).
A lot of spam and hate speech is automagically nuked at point of post, or soon after. 8-10 minute scrape cadences are going to miss a large portion of good deletes, strongly concentrating bad deletes.
There is no account for who deleted the comment; it's considered infeasible, but it's also said to be essential... Users deleting their own comments is censorship.
It's no secret that many communities uphold a comment threshold far above the legal speech level. You should expect to play by house rules. That's not censorship.
These seem like serious people, making a professional report, and they've put in a serious amount of work to code the comments they've seen, but every proportional argument is fatally flawed. Their source data is too flawed.
One of the issues that we have in the US is that there are (almost) no consequences for lying and hate speech. Politicians can (and do) make provably false statements with no consequence.
I don't believe "free speech" is a license to outright lie; so I pay attention to topics like this because I think it's very important to figure out how to control outright lies and manipulation in popular media.
And, a little closer to home, a few years back a person very close to me started reposting hate speech on Facebook. (It targeted a religious group, and was very reminiscent of Nazi anti-Jewish propaganda.) I don't know of a good way (in the US) to address that kind of hate speech, so it's really interesting to see how other countries try to reign it in.
46 comments
[ 2.5 ms ] story [ 86.4 ms ] threadWow, that's quite a bit worse than I would have guessed even as a critic of these types of laws.
This is an area where it's gonna be pretty subjective whether many comments are law breaking, so the internal guidelines for censors will likely err on the side of caution, which means removal.
This is exactly the worst hypocrisy of these very same megacorps, and particularly notable in mega tech corps, which have a particular tendency towards preaching loudly while practicing very differently.
On the one hand, they play themselves as arbiters of moral values and display a nearly (no scratch that, overtly, blatantly) preachy attitude towards their users and customers and even the wider public about social responsibility, fighting "incorrect" thinking or conduct, following proper regulations and etc and etc. With this attitude in mind, they essentially forcibly (at least within their reach as platforms) "guide" average people into certain types of behavior. They also make public speeches, blog posts, PR releases and all sorts of endless vomit about the de jure moral or social crusades of our time, depending on each company's specific context.
Aaand then, they also simultaneously perform nearly every kind of legal, social and financial malfeasance you can imagine, often blatantly for the sake of their real bottom line, which is market dominance and more financial power. (looking at just a few examples: OpenAI's blatantly predatory behavior toward pretty much everyone else's content, or its obviously dishonest crap about the supposed dangers of AI in its efforts to scare regulators in its favor, or Meta's too-numerous-to-count scandals across many years, or Google's intense quest to utterly destroy search for the sake of further ad dollars, or Amazon's indifference to an absolute avalanche of consumer fraud on its product sales platform while at the same time squeezing sellers as often and much as possible)
All of the above is to not even mention the completely presumptuous attitude among nearly all modern tech companies, and more recently, most other major consumer-oriented companies of all kinds (these days any large company is really a tech company at heart) that they have an absolute right to not only pry into everything you do no matter how private, but also store and distribute or sell that information as they please to nearly anyone at all with zero consequence. This last has been taken to the point where for all their talk of "valuing your privacy" the default behavior is the diametric opposite, and the default response to anyone who really tries to dial it back is overt hostility.
It's all quite grotesque.
How anyone could take seriously a single major major techcorp leader's preaching on any social issue is beyond me.
Says who? Right, the think tank.
How would you possibly evaluate such a claim seriously?
Companies drop posts for violations of community terms that have nothing to do with what is legal.
If you check their methodology, their coding guidelines were developed by legal experts who are native speakers in each of the target countries, so to answer "Says who?", lawyers.
[1] https://futurefreespeech.org/preventing-torrents-of-hate-or-...
[2]https://futurefreespeech.org/wp-content/uploads/2024/05/Prev...
The ad industry tends to be a lot more sensitive to certain content than governments are. (For example simple insults can easily get a Youtube video demonitized.)
There's a good chance that the removal of comments regarding killing and war is caused purely by ad regulations.
Without further information, these numbers tell nothing about the actual effects of censorship laws.
This is correct but they do show large national variation in the report.
> There's a good chance that the removal of comments regarding killing and war is caused purely by ad regulations.
This doesn't explain the Facebook deletions.
> Comments fell into this category if meaning could be interpreted from them, they were not illegal, did not contain linguistic attacks or fell outside the scope of the report (e.g. CSAM, revenge porn, marketing). Most comments belonged to this category.
https://futurefreespeech.org/wp-content/uploads/2024/05/Prev...
If you read further they say:
> Criminal restrictions on content that can be classed as The Future of Free Speech “revenge porn” or Child Sexual Abuse Material (CSAM) are discounted from discussion in this report as they are universally recognized as unprotected speech.
Section 2.3 Coding of Data:
> These legal notes are based on legislation related to freedom of expression, which means that violations of marketing legislation, for example, were not coded as illegal.
If I were to launch into a tirade about a particular race (pick one) if I phrased it reasonably enough to avoid falling afoul of hate speech laws it would be legal. It would also still be hate.
If I were to spread heinous lies about a group to inspire stochastic terrorism most places it would be legal.
Legality is a meaningless standard to apply to this situation.
The number of sites that limit moderation to strictly illegal material are few and far between, and they tend to degenerate into cesspools where only the worst elements prefer to gather. Even Gab recently learned this lesson[0,1].
[0] https://news.gab.com/2024/03/were-cleaning-up-gab-and-its-lo...
[1] https://news.ycombinator.com/item?id=39686704
I wouldn't trust any percentage in the report
You can distil the factual elements of the report to: "Comments that were available to us are now no longer available to us, but we don't know why"
> Does that mean they should get a pass and escape accountability?
What? My pointing out extreme flaws in data collection and analysis shouldn't mean anything other than there are problems with this report. Why are you trying to extrapolate a motive for me?
When was the last time the implications for the European Council was even mentioned in the context of (any) local election?
The problem is that there is no way to approve of a government locally, but disapprove of their performance in the EC; and given the two conflicting concerns, virtually everyone will favor the former over the latter. This means that de facto there is no voter accountability for the executive branch.
Having mere elements of democratic representation in a body that is empowered with overriding local laws and constitutions is not good enough.
Deputizing big tech to handle copyright disputes is in relative terms completely uncontroversial and is exactly the same means of censorship. Regulating the world's largest megaphones is the natural choice. It's hard to argue against any of the pieces when most people accept them individually but get uncomfortable at the reality of what it means put together.
I recognize it as a wart but I'm really not sure how to address it in an international context where there is no one frame of law that can be applied, nor should there be.
This whole article is idiotic. It can be re-phased as: "Meta and Alphabet don't want to hire large compliance teams, so they simply accept every request as valid." The companies have the option to push back, but they prefer not to, because they don't want to be bothered. And if a take-down request is not challenged, then we don't actually know if the courts would have found the speech illegal or not. If the courts frequently upheld the speech, then the regulations would naturally shrink in scope. But if the courts instead agreed that the speech violated the law, then the speech actually was illegal. But establishing the exact boundary between legal and illegal is something that can only be done if the companies, or some free-speech CSA, are willing to take some of these cases to court.
The criticism here is similar to the one made against content hosting companies who take down any content flagged by a DMCA request, even when the entity making the DMCA request is not the copyright holder. When the content hosting companies demonstrate cowardice regarding DMCA, then the DMCA can become a tool of abuse, but some of the criticism regarding the situation around the DMCA should be aimed at the cowardice of the companies.
It's worth repeating, too, Meta and Alphabet always have the option to leave the EU market if they find the regulations in the EU market to be overly burdensome. But if they stay in the EU market, then they'll have to abide by EU regulations.
(I claim that) Even without any censorship laws, Facebook and Youtube would remove lots of content discussing violent topics.
Without further information, we therefore cannot tell how heavily the censorship on these platforms is truly influenced by government law.
> Which is to say, it appears that discussing the news of the world is a good way to get censored in the European Union.
This article quickly draws heavy conclusions without providing the necessary evidence. It is very sensationalized.
There needs to be some sort of explanation for why there's so much variation between countries then.
There's a bit of handwaving about how the law is chilling free speech by encouraging companies to remove it too aggressively, but I don't think it's very convincing.
This is just another example of poorly targeted legislation with negative externalities that will far eclipse whatever problem is being solved.
The bigger problem is that we as a society for some godforsaken reason believe it’s best to change social behavior with law. This is the same shit we have been doing for centuries and it has never worked. You cannot stop people from internalizing being a human turd, although social pressures can. But if all you have is a hammer…
> Three actors have the ability to delete a comment: the user who posted it, the social media platform, or the owner of the channel or page, where the comment was made. This report does not have the capability to distinguish between these three scenarios. [..] Therefore, this report treats the three reasons for comment disappearance as a single phenomenon.
There's no connection at all with EU laws if they include the posts I delete myself.
> Comments fell into this category if meaning could be interpreted from them, they were not illegal, did not contain linguistic attacks or fell outside the scope of the report (e.g. CSAM, revenge porn, marketing).
If I'm reading this right, this contains content such as "Buy Coke!"
Is this useful data?
I certainly don't think it supports their stated position:
> In a world that works the way policymakers intend, we would expect to find that most deleted comments constitute illegal speech.
I'm not sure I agree with this position in the abstract either.
https://futurefreespeech.org/wp-content/uploads/2024/05/Prev...
But seriously, the data is scraped. In my experience, these networks are pretty finicky about which comments they'll actually show you (even if they're not deleted).
A lot of spam and hate speech is automagically nuked at point of post, or soon after. 8-10 minute scrape cadences are going to miss a large portion of good deletes, strongly concentrating bad deletes.
There is no account for who deleted the comment; it's considered infeasible, but it's also said to be essential... Users deleting their own comments is censorship.
It's no secret that many communities uphold a comment threshold far above the legal speech level. You should expect to play by house rules. That's not censorship.
These seem like serious people, making a professional report, and they've put in a serious amount of work to code the comments they've seen, but every proportional argument is fatally flawed. Their source data is too flawed.
I don't believe "free speech" is a license to outright lie; so I pay attention to topics like this because I think it's very important to figure out how to control outright lies and manipulation in popular media.
And, a little closer to home, a few years back a person very close to me started reposting hate speech on Facebook. (It targeted a religious group, and was very reminiscent of Nazi anti-Jewish propaganda.) I don't know of a good way (in the US) to address that kind of hate speech, so it's really interesting to see how other countries try to reign it in.