Commission opens formal proceedings against X under the Digital Services Act
ec.europa.euI’m curious if X will manage to build up the moderation tools and automation without rehiring as many moderators as they used to have. Looking at the requirements, some of those challenging - preventing fast dissemination of illegal content.
Having tried the "Added context" contributions, I think Twitter solved the moderation. It simply works. There's an abundance of free labor and it scales linearly with controversy/spread.
But that doesn't remove the content. It just sticks a label underneath it. And it's only on the most popular tweets.
There is a dude who lives a couple miles from me who flies a large confederate flag and has a statue of Robert E Lee in his yard. I am happy that authorities allow it because that way I know not to interact with him. I’d rather offensive people be loud and proud than secretive and using subterfuge to advance their cause.
That’s a feature, not a bug.
>It simply works.
Does it? Because I've seen it used to spread all sorts of bullshit, including against Musk/X.
One case that comes to my mind when Musk wanted to make blocking to work the same as muting, and the false claim of it being against Apple's App Store rules was added as context.
In what way does blocking not work the same as muting?
If you mute a user, you will not see posts from that user but they can see and interact with your posts. If you block a user, you will not see posts from that user and they cannot see or interact with your posts.
If the recent leak about the moderation policies for Twitter is correct, they don't actually want to remove that kind of content. So it might not be a matter of not having enough moderators right now but also an intentional policy decision to not sanction posts like this.
I can't find an English source here and the original is paywalled, but from a German article (https://www.heise.de/news/Leak-Leugnung-des-Holocaust-und-me...) summarizing the original reporting comes the following example of a tweet that is allegedly allowed now:
> Next stop on our tour across Poland is Auschwitz. For Jews this is the last stop, please exit and take your luggage with you.
It also mentions that threats of physical are no longer a reason for suspending accounts.
It's a pretty bad dark joke, but considering it illegal? Especially by an american company, freedom of speech and all that?
There is no absolute right of free speech in the EU (and probably most countries); foreign companies operating there are subject to local laws when doing so. (In my non-EU country that tweet could be interpreted as a hate crime too, quite uncontroversially)
Sure, but is this joke actually illegal?
I'm not as familiar with EU laws, but if I posted this joke online in the UK, then yes, it could easily be considered "grossly offensive" and therefore would be illegal.
I won't say it happens particularly often, but we do charge people in the UK for posting offensive jokes. Hell, you can be charged for saying something factually true if it's offensive in the UK.
If they are operating in the EU, it doesn't matter if it is an american company or not.
Sure, but can this joke itself be illegal? What law forbids a joke like this? It's not calling for violence, it's not denying the holocaust (just the opposite), it's just.. well.. a joke.
It’s absolutely inciting violence, just using coded language.
Even the EU is going to have some legal standards around this sort of thing, or you could spin whatever you want into being some sort of coded language for something that is illegal. I'm unfamiliar with EU (or Germany as it may be?) standards, but using the US as an example, there's the Miller Test [1] for determining obscenity. It has three parts:
---
1) Whether "the average person, applying contemporary community standards", would find that the work, taken as a whole, appeals to the prurient interest
2) Whether the work depicts or describes, in a patently offensive way, sexual conduct or excretory functions specifically defined by applicable state law
3) Whether the work, taken as a whole, lacks serious literary, artistic, political, or scientific value.
---
Something can only be deemed obscene if it meets all 3 criteria, which courts tend to interpret pretty rigidly. For instance, an average person is not the most sensitive or politically correct person. Of course it also somewhat amusingly emphasizes the cultural differences. Creative use of a nipple and some mayonnaise might be deemed as obscenity, while a graphic decapitation could not.
The Germans are not famous for making jokes about killing Jews, and their market: their rules.
Sure, but the germans did not outlaw jokes, even if they're germans.
It's not calling for violence, it's not denying the holocaust (quite the oposite), so why would it be illegal?
Bloody hell, that was an _example of a Tweet that they were okay with, in training material_?!
It isn't inciting violence, and despite what people say about Germans we haven't actually outlawed humor. The joke might be tasteless or insensitive, but I doubt it's illegal.
Of course many platforms wouldn't want such content on their platform, but there is also a vocal portion of people who have long called for platforms to behave more like town halls or public plazas, letting public discourse run uncensored as long as it isn't illegal. That's exactly what Musk seems to be doing
Oh, I’m not saying it’s illegal anywhere (maybe France? France is particularly aggressive on this sort of thing), but I would imagine that some Twitter salespeople will be having _interesting_ conversations with remaining large advertisers about this; it doesn’t really gel with previous positioning (at least to advertisers).
Musk didn't buy the company to make it gel with its previous positioning, for better or for worse. But I agree, it will be interesting if Twitter manages to retain any of the large advertisers, and if how they plan to become profitable again.
If you're critical of X allowing people to recite jokes about the holocaust, how do you feel about Netflix platforming similar jokes about genocides? (https://www.youtube.com/watch?v=CJurYs12ay4)
Do you feel some jokes are too offensive to be said? Do you feel the EU/UK are correct in their stance that offensive (aka hateful) jokes should be illegal?
"It was just a joke" isn't a legal defence when inciting hatred and violence. It's not legal even in the US (though IANAL).
Context matters.
The Supreme Court almost always judges in favor in free speech, including justices like Sotomayor. A place this frequently comes up is in rap where rappers frequently tend to rap about specific people and propose or fantasize about violent acts and even murder being committed against them. Eminem and his ex-wife Kim is a more or less well known example. And the Supreme Court has ruled, repeatedly, that it's 100% legal. A relevant court case is Elonis vs US [1]. He said some extremely explicit and provocative things, and the court ruled 8-1 in favor of throwing out his conviction. Interestingly enough, the sole dissenting vote was Clarence Thomas!
It is legal in the US as long as you are not intending and likely to incite imminent lawless action. Eg. To give a recent example, it's completely legal to chant "from the river to the sea, Palestine will be free" despite some interpreting that chant to be calling for genocide.
Yeah, that's why I'm saying context matters, and it's often up to the court to decide the context
Disappointing that the vast majority of top levels here are describing the EU as some sort of totalitarian police state. It does respect rights. Do the people in the USA think that every country should let its citizens act like uneducated armed griefers?
https://www.politico.eu/article/eu-commission-violation-priv... trying to ban end to end encryption is the opposite of respecting rights.
Chat control was struck down last month and was never going to go through anyways as just having a cursory understanding of technology is enough to understand that banning end to end encryption is infeasible.
And chat control being struck down was in part on the grounds of respecting rights around private correspondence so you're just spreading factually wrong information for no good reason.
The fact that it got as far as it did is incredibly problematic and does not paint the EU as an organization that respects rights.
There's the commission and the parliament. The commission comes up with ideas for laws and the parliament says if they're okay or not.
If your government is working properly, there should be a small part of it trying to ban encryption, and a part of it trying to keep encryption legal. It's an adversarial system. Encryption will be an unpopular subject here, so let's talk about zoning instead: a small part of the government should be representing builders by trying to make it legal to build anything anywhere, and a small part should be representing land speculators by trying to make it illegal to build anything anywhere.
>EU as some sort of totalitarian police state
Because it (or parts of it) were recently trying to ban private online communication, which is exactly what a totalitarian police state does.
To people downvoting the above, you should inform yourselves better: https://www.politico.eu/article/eu-commission-violation-priv... . The EU Commission has been trying to ban end-to-end encryption, which is functionally equivalent to a ban on private online communication.
Interesting that it is timed with the recent launch of Threads in EU. Will definitely establish a compliance benchmark for X.
But the big problem for X is that Musk burned so many bridges with fired employees that the list of whistleblowers for any issues will be endless.
I think the timing is _probably_ coincidental, more down to a bunch of different things happening in European law currently than anything else. The Threads delay seems to be down to Digital Markets Act compliance tweaks (the DMA entered into force earlier this year); this is under the Digital Services Act, which enters into force in two weeks.
> These are the first formal proceedings launched by the Commission to enforce the first EU-wide horizontal framework for online platforms' responsibility, just 3 years from its proposal.
A mere 3 years! This is, in fairness, quite proactive by EC standards.
>A mere 3 years! This is, in fairness, quite proactive by EC standards.
Isn't that just because EU rules have quite long transition periods to give companies time to comply with the new rules? GDPR had a two-year transition period before enforcement, and even then there was tons of complaints towards the end of the two-year period on how there just wasn't enough time to comply.
Well, there are many things that make the EC have a reputation of being slow.
One thing is just the transition periods you mention, then there is the delays caused by the need to have legislation at the EU levels turned into actual national legislation to be implemented.
Then there is the whole trialogue thing where, after first approving a draft law, the EP and the EC have to hold further negotiations on a common text while supervised by the Council, which tended to happen behind closed doors so the experience is that the contentious law that was just approved is quietly sucked into a black hole and then you might hear of it being passed only many many months later.
Finally, we also have instances like with the Chapter 7 investigations against Orban where the commission knows the council is likely to block any conclusions so there's no point in rapidly pushing ahead.
Even after that transition period, _serious_ enforcement didn’t start for years. Prior to 2021, the largest fine under the GDPR had been 50 million (in 2021-2023 the Irish regulator lurched into actual and fined Facebook a couple billion).
I suspect the “local regulators do enforcement” model won’t feature heavily in future EU law; it doesn’t work very well.
Commissioner Breton sent a stark warning to Musk, that was arrogantly brushed off in typical Musk fashion. Musk is now going to be schooled in very expensive fashion as to how being a billionaire does not make you above the laws, in Europe at least, and this is probably why the proceedings were launched, although in all likelihood the EC does not have all its ducks in a row in terms of organization, staffing and processes to enforce the DSA and DMA just yet.
Which law(s) did they break?
And about time. The EU (and the US even more so) has been pretty ignorant about the way hostile anti-democratic regimes wage war against free and democratic societies.
For dictatorships a misinformation campaign is both way more effective and cheaper than a single Mig fighter jet. This asymmetric war was already a big problem, but the oligarch Musk might have been too loud a siren for the sleepy lawmakers that believed that democracies will survive no matter how much you are killing its base.
- yes, there are all kind of problems people are rightly upset about.
- yes, politicians have traded trust for political results.
- yes, the US/EU has also failed by opting for "Real Politik" instead of values.
- no, a whattabouttism doesn´t help our societies. Instead, fight to preserve and improve what we inherited.
- no, helping to fuel distrust undercuts the fabric of our societies.
Never thought a plea to save our democracies by making a distinction between honest criticism one the hand, and sponsored covid disinformation campaigns on the other would attract this response.
Ones wrong opinions are absolutely tolerated. We all have them. Hybrid warfare is something else and should not be.
Disinformation is not harmless: half of the US believes that the elections were stolen. If people don´t believe in democracy anymore, it is game over without people realizing it.
> But the big problem for X is that Musk burned so many bridges with fired employees that the list of whistleblowers for any issues will be endless.
That is great, I don't think any company should rely on employee trust to keep illegal practices under wraps. If you put it like this maybe it would be good for Meta to piss off their employees also so we can also get some insight into their illegal practices.
How will they whistleblow against him if he fired them in the first week before he did anything?
Because he didn't fire everyone in the first week.
Not flagging because it is more interesting than usual for a link about Twitter:
it's the first time the law is used, and it also _strongly_ hint that they tried first to get access to the informations without legal proceedings.
The legalese is strong, but I think most of it is understandable (which is surprising).
But: Twitter is now like the 12th social network, who cares?
I think the interesting bit here is how quickly they’re going out of the gate, really; the DSA doesn’t even come into force for another two weeks. Companies who were depending on DMA and DSA enforcement being on the same basis as GDPR enforcement (ie extremely slow) may be disappointed.
The EU learned from GDPR, that member countries cannot be relied to enforce the law when some like Ireland have deliberately positioned themselves as a tax (and now GDPR) shelter for multinationals. That's why DSA and DMA enforcement is centralized in Brussels.
Just make a new rule that you can only post content on x that has first been posted on a properly moderated social media site.
Musk keeps repeating that X always follows local laws and regulations, but in practice it seems to only apply to countries like Turkey and Saudi Arabia. Curious.
Interesting that Community Notes is mentioned, because it's by far the most reliable anti-"disinformation" mechanism I've seen.
The "bluecheck" system has been absolutely wrecked, though. Even with the haphazard patching of grey and gold ticks, a bluecheck more often than not is the sign of self-promoted nonsense.
Yep, the great thing about community notes is, that it leaves the tweet unchanged and just adds context (or refutal, or whatever). It turns censorship into (hopefully) "argumented refutal".
Did people actually rely on the bluechecks anyway in the first place? For me it was at best a slight help if I was trying to find a celebrity that had impersonators. I would be surprised if I’m a very typical user of course. But still I had the impression the bluecheck system was fondly replaced.
> For me it was at best a slight help if I was trying to find a celebrity that had impersonators.
Yes, that is literally what it was designed for, and the only thing it was useful for.
It’s now, I suppose, useful for identifying people to avoid, but it’s no longer useful for preventing impersonation.
> For me it was at best a slight help if I was trying to find a celebrity that had impersonators.
That's exactly why it was useful. Major institutions and public figures have impersonators, some satirical and others intentionally misleading. The check made it easy to sift what actually came from someone and what was either lies or satire.
On that front it hasn't gotten too bad because the grey and gold checkmarks exist.
Now blue checkmarks are a sign to ignore the tweet because it's someone who's just paid to boost their signal. Of course this has rendered the thread of replies to a given tweet unusable since the top posts are all posts by people who have paid to be seen rather than posts with more likes or retweets (actual signals of value/popularity).
It was actually a great system. All those Elon free crypto giveaway scam accounts could make a profile that looked exactly like Elon Musk, but they couldn't get that blue check. It was great when it was a completely neutral system to state only "this account is who it says it is." But then they ruined it by removing the checks from people with unpopular political opinions, even though they were in fact the official account of that person, and then it became seen as a sort of political endorsement from Twitter.
I know of one example where community notes was used to spread outright lies about the victim of police brutality a full day after the claims were disproven, so it still needs some work.
Another amusing thing right now, X marked a journalistic article about Tesla as "spam, misleading or violent":
https://www.expressen.se/nyheter/sverige/x-marker-lank-till-...
I don't understand why Elon hasn't hijacked Community Notes for his own purposes yet.
Of course they will come after that. Because when the government says "remove disinformation" they mean "remove things that we don't like," and community notes doesn't do that.
"news" means something "new is happening", how do you know if it is a disinformation or right when it just came up?
Should we blindly trust and consume news from mass media propaganda machines like CNN, Fox news, BBC and etc?
Liberalism was meant to end this kind of EU censorship. Parliaments were supposed to be the place where diverse and illiberal speech was dissolved into a new liberal language, not through brain dead regulation and technocratism.
If Meta properties are the example of what to expect from the future locked-down internet, prepare for the internet and much more to wither on the vine and die a slow and sad death.
> If Meta properties are the example of what to expect
Then you see why governments end up regulating them. Not because "liberalism was meant to do something magical"
"among others, concerned the dissemination of illegal content in the context of Hamas' terrorist attacks against Israel, the Commission has decided to open formal infringement proceedings against X under the Digital Services Act."
Why not open a case against Israel and some of the citizens for spreading a misinformation about beheaded babies and some other lies?
So what's the betting that X will be unavailable in Europe by end of day?
I will take the counter bet, and even go as far as saying X will remain available in EU for the next month at least.
Oh, this procedure will be running for years. Expect a few rounds of "we're not complying / we are complying / oh no you're not" as well, each of which taking over a year.
Because X will have gone bankrupt and bust worldwide by then?
Nope, because I don't think X is going anywhere anytime soon, even though its demise is reported as inevitable at least every week.
Don't threaten me with a good time.
Musk will run his mouth and then implement everything required.
EU just wants to regulate everything. Recent ai regulations are very bad. Purposefully nerfing yourself at the technology front isn't a solution.
>Purposefully nerfing yourself at the technology front isn't a solution
It is a solution if the problem is "preventing anything that could challenge the entrenched elites".
It doesn't have to anything with elites. EU just likes to virtue signal on climate change and new technologies by introducing regulations to keep status quo. Keeping jobs for the sake of only keeping jobs doesn't make any productive sense.
The EU government doesn't give a crap about unfair social media algorithms when they benefit their authoritarian agendas (at the expense of citizens) but as soon as soon as the algorithms go against their own agenda, they immediately launch a commission! And, ironically, the algorithms today are less constrained than they've ever been.
You realize that the European Commission[1,2] is a big part of the executive branch of the EU, yeah? (The other one being the european council). So if there is something as an "EU government", it is the commission.
[1] https://commission.europa.eu/index_en [2] https://en.wikipedia.org/wiki/European_Commission
The European Commission is not elected by EU citizens btw.
I know. The same holds true for the German government.
What authoritarian agenda? What commission?
Maybe they saw something on youtube that said Von Der Leyen is going to the ISS? That's about the only way I can see "EU launches comission" make any sense. ;)
For example, they were happy about social media platforms censoring or otherwise suppressing content which opposed the COVID19 vaccine mandates; they didn't raise any alarms there while there was actual substantial manipulation of the algorithms involved...
But now that X has mostly removed filtering and manipulations from the algorithms, they begin proceedings against it! Just because they don't like the kind of content that people posted. They don't want people to have equal voice.
Not sure what the issue is? European countries not being allowed to have their own culture around speech (which rolls up into the EU to some extent)?
I don't think it's the job of EU government to manipulate the algorithms to suit their values. Their job is the opposite of that; to ensure that the algorithms operate with as little manipulation as possible so that every individual has equal voice/reach on these platforms (at least equal on a per-follower basis) with as little manipulation as possible.
As someone who didn't have much prejudice about the current Israel/Palestine situation, I appreciated seeing both sides; it allowed me to come to the conclusion that both sides fall short of my own moral values and therefore I don't need to involve myself in debates about the lesser or two evils.
Maybe other people with different morals than myself felt that one side met their 'good guy' threshold and wants to support them. They're free to voice their support, but I'd like to keep my taxpayer money in my own country where it can support people whose morals more closely align with my own.
The EU as well as nation states have laws to uphold and policy/social objectives that might require that. The way to change that is to change laws, customs, and objectives. You are welcome to push down that route, of course.
> You are welcome to push down that route, of course.
Of course not, precisely because public speech including on media such as twitter is heavily censored.
The trick is gross but effective: treat ideas threatening to the system as illegal so they can be fought using the legal system instead of the political one. It's a power grab to silence Europeans. Same old trick used in China, Iran, any autocracy really.
There is nothing heavily censored. You have new parties rising up, for example, nothing stopping that.
> to ensure that the algorithms operate with as little manipulation as possible so that every individual has equal voice/reach on these platforms
The companies do that? In which alternate world do you live?
Please realize that the anti-vaccine nonsense is content driven by hostile regimes, ie China and Russia.
I will give you some links:
- https://www.rferl.org/a/russia-china-covid-disinformation-ca...
- https://www.bmj.com/company/newsroom/foreign-disinformation-...
- https://www.politico.eu/article/antivax-conspiracy-lean-pro-...
Who said anything about vaccines? Some European countries have been penalising people for entirely organic anti-immigration sentiments.
I have never heard of people being "penalized". If you violate the law, for example by discriminating someone, you can get a penalty indeed.
You can have all your "anti-immigration sentiments", but you are not going to be penalized. You can be completely opposed against immigration, and if you have some good arguments, you can calmly post them anywhere. I fear that "sentiment" is doing a lot of work here.
Besides: what looks like organic in extremist circles is often not. You start with a business model and you will gather a herd. You just have to feed them.
European countries are under attack by the Russian regime, if you missed it. Immigration was and is used as a weapon to destabilize European countries. Russia was/is financing immigration hostile parties and organizations in Europe.
An example:
https://www.dw.com/en/poland-says-belarus-russia-behind-new-...
https://www.lemonde.fr/en/france/article/2023/09/22/le-pen-s...
Because of the Russian war against the Ukraine, Europe currently hosts millions (> 4 million) of additional refugees.
Exactly, "the effectiveness of measures taken to combat information manipulation" sure didn't seem to be an issue for decades for the UE commission when thr manipulation was done by a cell at Twitter itself, such as removing trending conversative hashtags or suspending right-wing EU politicians.
Of course now that the platform allows more balanced opinions, any power available will be used to silence (growing) opposition.
Just pull out of the EU. It's overtly hostile to freedom of speech. Run Twitter from the US, and let the EU erect their own "Great Firewall" if they want to block it. F500s (or at least the ones still advertising) can pay for ads through their US subsidiaries, and the handful of remaining Twitter employees in Europe can either take a relocation package or severance.
How is fighting against dark patterns, opaque and arbitrary moderation decisions (among other things) an attack on (your very american conception of) free speech?
They seem largely focused on "illegal content", specifically referencing the Oct 7 attack on Israel:
> concerned the dissemination of illegal content in the context of Hamas' terrorist attacks against Israel
> The compliance with the DSA obligations related to countering the dissemination of illegal content in the EU, notably in relation to the risk assessment and mitigation measures adopted by X to counter the dissemination of illegal content in the EU, as well as the functioning of the notice and action mechanism for illegal content in the EU mandated by the DSA, including in light of X's content moderation resources.
At least in the US, footage of a terrorist attack is not illegal. In fact, it often serves important journalistic and public interest purposes. Twitter is an important source of first-hand coverage of conflicts such as the Israel/Hamas conflict and the Russian invasion of Ukraine. EU safetyism threatens to interfere with this important function.
Part of the German government is trying to make it illegal to say bad things about Israel.
"freedom of speech" is just Virtue signalling.
Not sure why would Twitter leave EU - Musk's Twitter has no problem suppressing freedom of speech, like in that case of Turkey.