According to the imaginatory rules of lawyers / bankers in power.

Most countries have laws against foreign interference. I presume disproportionately propagating far-right content via algorithm might very well count as such an interference.

How do you prove it?

Not that hard really (found only German example though)

https://www.linkedin.com/posts/global-witness_globalwitnessi...

That doesn't really prove that much. They only made 3 accounts on X, then with each account they followed the 4 major parties and their leaders. Not a very good sample size.

Additionally, the underlying algo could lean toward to what causes more engagement rather than equally tailoring each party. Without having any information personally about German politics or their cultural engagement with social media, I can only anecdotally say that in America, it appears to me that the ones most engaged with social media on X are more often than not right leaning. More of the left-leaning engagement seems increasingly moving towards other outlets, such as Bluesky.

Social media is a populist's dream come true.

All the anger is amplified through an algorithm and you can't even say it's the algorithm's fault.

> I can only anecdotally say that in America, it appears to me that the ones most engaged with social media on X are more often than not right leaning. More of the left-leaning engagement seems increasingly moving towards other outlets, such as Bluesky.

Isn't that movement because Musk changed X to increase promotion of right-wing things and decrease promotion of left-wing things?

Before Musk Twitter did promote right-wing stuff more than left-wing stuff. I don't have a cite but there was a paper published by researchers who had been given full access to internal Twitter data that showed this.

That level of right-leaning is probably the level of right-leaning you get from the natural level of engagement you get on a fairly neutral platform from right-wing people being more likely to engage than left-wing people.

After Elon bought Twitter, they ended up losing something like 10% of their users. That's going to be made up mostly of highly ideological ostensibly left-leaning Americans, likely with a much higher than average "engagement." And the lion's share of that happened long before there were any meaningful algorithmic changes. That alone is going to create a meaningful shift in the status quo on overtly political topics.

> That doesn't really prove that much.

You seem confused. France launched a criminal probe, whose goal is to investigate whether these types of biases reflect actual manipulation from Twitter. You are presented with evidence suggesting these biases exist. Now France is expected to look into them. Their findings will either support a criminal case, reject a criminal case, or simply not support further investigations.

> Pushing antisemitism, holocaust denial, and supremacist content is not ok, even if you argue they are only pushing extremist content for the clicks (which is a laughably bad hypothesis as Twitter's business model depends on the platform having mainstream appeal)

this is the playbook. they dont need to prove it beyond some weak deduction mecanism. they did it for romania to cancel an election

What if e.g. right-leaning content gets more likes/upvotes and this makes the algorithm show the posts more? That way, the algorithm can still be neutral, but the result is not.

"The left can't meme" is just one instance of so-called "meme magic" that came good, if good is what you see in the enterprise or if the particular ends justified the memes.

https://knowyourmeme.com/memes/the-left-cant-meme

https://knowyourmeme.com/memes/meme-magic

https://unherd.com/2021/08/why-the-left-cant-meme/ | https://web.archive.org/web/20210924231415/https://unherd.co...

https://www.vice.com/en/article/trumps-occult-online-support... | https://web.archive.org/web/20170125155034/http://motherboar...

As for me? I am beside myself. I view the content as part performance, part farce, all kayfabe. Trump et al learned from the best:

https://en.wikipedia.org/wiki/HyperNormalisation#A_World_Wit...

> In Russia, Vladimir Putin and his cabinet of political technologists create mass confusion. Vladislav Surkov uses ideas from art to turn Russian politics into a bewildering piece of theatre. Donald Trump employs the same techniques in his presidential campaign by using language from Occupy Wall Street. Curtis asserts that Trump "defeated journalism" by rendering its fact-checking abilities irrelevant.

> The American Left's attempt to resist Trump on the internet had no effect. In fact, they were just feeding the social media corporations who valued their many additional clicks.

https://en.wikipedia.org/wiki/Vladislav_Surkov

> Vladislav Yuryevich Surkov (Russian: Владислав Юрьевич Сурков; born 21 September 1964) is a Russian politician. He was First Deputy Chief of the Russian Presidential Administration from 1999 to 2011, during which time he was often viewed as the main ideologist of the Kremlin who proposed and implemented the concept of sovereign democracy in Russia. From December 2011 until May 2013, Surkov served as the Russian Federation's Deputy Prime Minister. After his resignation, Surkov returned to the Presidential Executive Office and became a personal adviser of Vladimir Putin on relationships with Abkhazia, South Ossetia and Ukraine.

> BBC documentary filmmaker Adam Curtis credits Surkov's blend of theater and politics with keeping Putin, and Putin's chosen successors, in power since 2000.

https://en.wikipedia.org/wiki/Sovereign_democracy

> Whilst talking about sovereign democracy in 2006, Mikhail Kasyanov said that "... the aims of this doctrine are quite clear: the concentration and holding of political power and property at any cost. The consequences of this are already evident, including the glorification of populism, the steady destruction of private and public institutions and the departure from the principles of the law, democracy, and the free market."

> United States Assistant Secretary of State for European and Eurasian Affairs Daniel Fried (in office 2005–2009) stated in a 2007 interview:

> > I get nervous when people put labels in front of democracy. Sovereign democracy, managed democracy, people's democracy, socialist democracy, Aryan democracy, Islamic democracy—I am not a big fan of adjectives. Managed democracy doesn't sound like democracy. Sovereign democracy strikes me as meaningless.

https://en.wikipedia.org/wiki/Guided_democracy

> Guided democracy, also called directed democracy and managed democracy, is a formally democratic government that functions as a de facto authoritarian government or, in some cases, as an autocratic government. Such hybrid regimes are legitimized by elections, but do not change the state's policies, motives, and goals.

> In a guided democracy, the government controls elections such that the people can exercise democratic rights without truly changing public policy. While they follow basic democratic principles, there can be major deviations towards authoritarianism. Under managed democracy, the state's continuous use of propaganda techniques, such as through manufacturing consent, prevents the electorate from having a significant impact on policy.

> The concept is also related to semi-democracy, also known as anocracy.

https://en.wikipedia.org/wiki/I%27ve_Got_No_Strings

> Down where the Volga flows

> There's a Russian rendez-vous

> Where me and Ivan go

> But I'd rather go with you, hey

> There are no strings on me

Be well, anon.

Subpoena the company and demand registration data for accounts flagged as looking like foreign influence. If it's a foreign IP or email, it's foreign influence. If it's through a proxy or VPN, it's likely foreign influence.

It is that simple. If Twitter doesn't comply with court orders like that, they will no longer be allowed to operate within France or the EU. Never ever think companies are above the law or don't have to cooperate with requests like this, that's a defeatist attitude.

That's literally a fishing expedition that could be exploited to harvest information on whoever one likes. It's the reason our legal system requires a substantial burden of evidence before one can even think of pursuing a warrant.

I mean seriously think about the implications of your concept here. X is a worldwide service. Should X just be an information broker dealing out the details on users to any country that makes claims as nebulous as 'foreign influence'? It's a term which can be (and often is) whimsically applied to essentially anybody who happens to disagree with some regime in a given country.

"Should X just be an information broker dealing out the details on users to any country that makes claims as nebulous as 'foreign influence'?"

Yes. A poster that specifically tries to influence a French elections with false statements is breaking French law wherever and whoever they are.

So should X give information to Iran on any Jewish account it suspects of engaging in "foreign influence"? How about to Russia on any Ukraine account, or vice versa? What about to everywhere else in the world and information on US accounts some country somewhere suspects of engaging in "foreign influence"?

Somehow rules seem to no longer apply when we start consider them being applied against our interests. This is a big part of the reason why the 'rules based order' is now mostly just sardonic mockery.

X is already doing it with India and many other countries, so your argument is nul.

https://www.washingtonpost.com/world/2023/11/08/india-twitte...

The situations seem unrelated. There's a much more informative article on what's happening in India here. [1] India is censoring accounts in India which X complies with by making it clear that's what is happening as well as providing information to the user on how to legally challenge it.

France, by contrast, seems to be making poorly defined and largely unfalsifiable claims about X's algorithm itself. I do not believe X would comply if a country began demanding algorithmic changes, as the main reason you'd go after the algorithm is to enact censorship with no transparency, which is very much in the spirit of contemporary French politics.

[1] - https://www.hindustantimes.com/india-news/elon-musks-x-says-...

Of course it is falsifiable, that is why they ask to review what is the algorithm, what was changed and why. Given how mercurial Musk is, you can be sure that he's asked his team to bias the algo - not least to boost himself. Again, perfectly demonstrable in a factual way.

X already complies with Germany's ban on nazi material. If that's not an lagorithmic change, what is?

>which is very much in the spirit of contemporary French politics.

Put your own (glass)house in order. https://www.theguardian.com/us-news/2025/jun/19/us-free-spee...

You're again describing something where there is straight forward censorship. This is relatively easily done without algorithmic changes and enables X to provide complete transparency to the user about the censorship, as well as any legal recourse or other options a user might have. France, instead, wants the algorithm to promote material in a certain way which is radically different, and can be utilized to censor things without any accountability or transparency.

As for falsifiability, proving a negative is generally impossible. This is how witch trials were able to be successful. You can't really prove you aren't a witch, when the person accusing you already assumes that you are. This is why the burden of proof is on the person making accusations, as opposed to 'Let me go into your house and see if I can find anything that I might be able to use as evidence to justify my hunch that you're a witch. Ah hah, I found a dead frog! Burn the witch!'

>can be utilized to censor things without any accountability or transparency

What is the current accountability or transparency of the X algorithm?

Ah yes, Musk promised to make it public, and... it hasn't been updated in two years. https://github.com/twitter/the-algorithm And clearly, the promotion of Musk's tweets on everyone's timeline is completely organic.

>This is why the burden of proof is on the person making accusations Yeah, this is all a WITCH HUNT! No justice department has ever done any investigation in the history of humanity, warrants don't exist, people are only judged by what is publicly known without police ever seizing any private material. WITCH HUNT!

Go back to truth social, mister President.

There is obviously no accountability for social media algorithms, anywhere. Facebook/Instagram/YouTube are conspicuously algorithmically downranking (and on occasion overtly censoring) pro-Palestinian content, but the political establishment in Western bloc nations don't care, because they don't want unbiased algorithms - they want algorithms that are biased to push their issues of the day.

And Elon is being followed by some 220 million+ people. That's a not insubstantial chunk of the entire human race, let alone the subset of X users. It's unsurprising that his comments would end up highly promoted by most of any algorithm that significantly weights account popularity, which all do - excepting perhaps TikTok.

To get a warrant you need to present strong probable cause to a judge - not liking what is being promoted would not suffice.

There are plenty of investigations regarding this, for example: https://www.justice.gov/archives/opa/pr/justice-department-l...

Have you already forgotten the US want to fully ban TikTok?

https://en.wikipedia.org/wiki/Restrictions_on_TikTok_in_the_...

Define disproportionately propagating far-right content.

Where is Mecha Hitler when you need him?

> Define disproportionately propagating far-right content.

How about:

> Elon Musk’s artificial intelligence firm xAI has deleted “inappropriate” posts on X after the company’s chatbot, Grok, began praising Adolf Hitler, referring to itself as MechaHitler and making antisemitic comments in response to user queries.

[…]

> “The white man stands for innovation, grit and not bending to PC nonsense,” Grok said in a subsequent post.

* https://www.theguardian.com/technology/2025/jul/09/grok-ai-p...

Or:

> In a series of posts – often picking up language from users or responding to their goading – Grok repeatedly abused [Polish PM] Tusk as “a fucking traitor”, “a ginger whore” and said the former European Council president was “an opportunist who sells sovereignty for EU jobs”.

* https://www.theguardian.com/technology/2025/jul/08/musks-gro...

Or last month's

> Elon Musk’s artificial intelligence chatbot Grok had been repeatedly mentioning “white genocide” in South Africa in its responses to unrelated topics and telling users it was “instructed by my creators” to accept the genocide “as real and racially motivated”.

* https://www.theguardian.com/technology/2025/may/14/elon-musk...

These things kept happening for 2-3 years with pre-ChatGPT LLM tech demos from Microsoft, Meta and Google.

Not to be an apologist of course, just wanted to point out that it may be a sign of them being behind the curve on some technical aspects, or at least on best practice, likely on purpose. Sure they probably did some ideological meddling too.

For all the (valid) criticism on alignment/censoring, one has to acknowledge the success of the pragmatic approach from OpenAI and Anthropic. As much as we might not want to admit it, bit of censoring is kind of critical to be able to use LLMs seriously to solve real problems.

A bot went off the rails, and was subsequently corrected. Again, what is the problem? I thought France had freedom of speech but maybe not?

Friendly reminder that

> In France, a woman spent 23 hours in custody for giving French President Emmanuel Macron the middle finger. (She was acquitted after arguing she had pointed her finger in the air and not directly at the president.)

https://archive.is/MNgc6

Friendly reminder too: https://www.bbc.com/news/world-us-canada-41892544

And far worse is happening right now against law firms and media companies.

> I thought France had freedom of speech but maybe not?

Freedom of speech in France (and many other countries) is not the same as the US (assuming you are from the US). Apart from the question of whether it even applies for an AI, it does not protect hate speech.

Sorry if you ban speech to the right of center then you don't have freedom of speech. "hate" is a loaded term, poorly defined. To me, it might seem you hate everyone right of center, while you think there's some clear line where speech crosses such that it can be labeled far right and a ministry of truth can slap a fine down to halt this "dangerous" speech. Meanwhile, is the "danger" equally regulated across the political spectrum and covering far-left speech as well? Of course OP didn't even bother to balance the concern about far right speech with equal concern about far left speech. You can be certain that such biases exist in any institution in which you might construct to attempt to regulate and that it would be absolutely impossible to obtain a perfect and permanent balance between the factions. Inevitably the balance of power would shift in favor of one faction or another and the more likely outcome is that the faction would use its monopoly over "truth" or "hate" to redefine even valid and factual criticism to be hate speech. Perhaps your country has already hit that tipping point and you can't even save it because you lack the liberty to do so?

God how I hate these sorts of "far-right" VS "far-left" points.

First, the far-left is _nowhere near_ as far as the far-right is (in fact, I don't think there's _any_ significant far-left representation in France: neither the Communist Party nor the France Insoumise are far-left, they're just left).

Second, you make it sound like there's a bias towards the left, which is a ridiculous point that has no ground in reality: France is _heavily_ biased towards the right. The internal and international politics are deeply right-wing and getting _more_ right-wing by the year, most successful politicians are on the right, even the so-called left-wing governments have been doing centrist or right-wing politics since 1995: the left is _weak_ in France despite popular support. So "your country has already hit that tipping point and you can't even save it because you lack the liberty to do so" is laughable: the _people in power_ are regularly indulging in hate speech.

Third, you make it sound like the hate speech law is biased to protect the left, when the reality is that hate speech almost exclusively exists on the far-right so _obviously_ it's going to affect the right more. Let's talk about concrete examples:

You'll often hear from the far-right that people of colour are fundamentally inferior to white people. That someone of a particular ethnicity is a "savage", or is uncultured. That muslims are terrorists, jews are conspiring. There's so many examples. That's all horrific hate speech that's forbidden in France. Please do give concrete examples of far-left hate speech that you hear these days that is remotely comparable, because I cannot think of any. Maybe "kill all billionaires"? Which by the way _is_ illegal, it's incitation to violence.

So yeah, some speech at the far right (not "right of center") is forbidden in France. You seem to think that's a limitation of freedom, but the minorities feel _way_ more free for it (although they're still not having a great time, hate speech still happens).

Your country is demanding access to all social media of incoming visitors, jailing students for alleged antisemitism, putting hundreds on administrative leave for signing a letter about how cuts are making their job impossible, shaking money off the pockets of media companies and law firms under threat of retribution, and using medical records to arrest undocumented immigrants.

I'd see two sides at least:

- the bot can't work in a legal void. It's pushing messages on a public platform , someone has the responsibility for that.

- if correcting speech is enough we should all be free push whatever horrible things we want online, subsequently correct them and never face any consequences whatsoever.

> A bot went off the rails, and was subsequently corrected.

How sure are you about that? I mean, it is more likely that the bot overshot how it was expected to push propaganda. Musk is on record expressing disagreement with how LLMs are trained to be "politicaly-correct", which is a dog whistle for pushing extremist views.

> Again, what is the problem? I thought France had freedom of speech but maybe not?

This is a puerile and simplistic view of what freedom of speech is. You are free t speak your mind, but others around you are free from experiencing abuse and discrimination. Also, media has more responsibilities than morons running their mouth, and even those are liable for hate speech.

> You are free t speak your mind, but others around you are free from experiencing abuse and discrimination.

No, you can't have freedom of speech and forbid the of hurting someone's feelings. The very speech that needs to be protected is damningly truthful speech. I'll suffer through a million fools and idiots spouting off nonsense just so that one person who, against all popular delusion and immense social pressure, when men and women in large numbers often go mad with hysteria as if a viral fad, and with the fortitude of knowing they are correct, can appeal to society at large without fear of arrest or legal consequences. These are the ones who change history for the better. The fools are like gnats, they are quite easy to ignore and swat off. The barriers you construct to make people stay within the lines are shackles on thought itself. These barriers inevitably get reinforced by the powerful until no thinking is done at all.

France's freedom of speech is as strong, or stronger, than the US.

It’s different to the US in some important ways.

There are strict limitations on 'hate speech’, denial of the Holocaust is illegal, and there are laws still on the books (and some examples of media outlets being prosecuted for breaking those laws) around presenting drug use as a positive thing, or encouraging drug use.

You can be prosecuted as an "apologist for terrorism” should the government conclude that this is what you are doing. You can also be charged with “contempt of public officials” as people were for burning an effigy of President Macron.

In the US, as far as I know as someone who has only ever visited the country for a short time, you are allowed to hold the President in contempt, you can announce to anyone who’ll listen your ignorant, racist, Holocaust-denying opinions and not be afraid of that speech being criminalized (though there are social costs you’ll probably pay), and if you want to go on the internet and encourage people to try drugs, you can. You can support whatever side of whatever conflict around the world you like with your words and you probably won’t be breaking the law.

France’s laws around freedom of expression are strong, but they are different to those in the US and I would say offer fewer protections for citizens than the US 1st amendnment.

I think I'm the US, those are all things you could do until recently. 1st amendment is being more and more ignored currently

In the US and most countries though, there's libel / defamation laws; you can say a lot of shit about people, but they can then sue you for libel / defamation if you spout nonsense.

Those same laws exist here in France, too, even if they are a little more complex. The truth of a statement, for example, is not a sufficient defence if that truth is not something that is public. Making someones private life public, even if your comment is true, can be against the law.

Defamation might be illegal in both the US and France, but I can burn an American flag and show contempt for the US President without commiting a crime the US. Do the same in the France and you don’t have the same protections for your speech.

Libel and defamation open you to civil suits but only perjury is criminal.

Criminal defamation is rarely prosecuted, but it's on the books in a large fraction of US states: https://en.wikipedia.org/wiki/United_States_defamation_law#C...

If you don't see the problem with far-right rhetoric then you are part of the problem. Europe has free speech protections, but only within reason; nazi symbolism is explicitly banned in Germany for example, and holocaust denial is also illegal.

Then there's some things you can say without getting into legal trouble, but you may get killed for it by those you offend.

No I don't see the problem. We've had Freedom of Speech for 250 years in the USA and our democracy has lasted longer than anyone's. France is on their 15th or 16th government in that amount of time. Perhaps if the copied the Bill of Rights verbatim they would still be on their 1st.

Don't even get me started on Nazis. Their most effective tool for control over citizens was restriction of speech. The Reichstag Fire decree being exhibit A.

The far left is just as awful and the USSR's glavlit, gulag, etc were extremely effective at silencing criticism. All ever in the name of order and peace, but in truth a consolidation of power.

> Then there's some things you can say without getting into legal trouble, but you may get killed for it by those you offend.

Sure, this is why Americans have a phrase called "Live free or die". If you're not willing to die for liberty then your fear of death will be used to control you. So fuck em, say what you think.

I would say after the bot has gone off the rail three+ times and conveniently pushed specific agendas (or just went full racist) that it's probably time for an investigation to figure out who's doing what and why.

We have freedom of movement but there are retricted areas.

We have freedom of religion but you can't e.g. declare "child porn" is your religion.

We have freedom of assembly but the police can disperse a crowd.

And that's the case everywhere, the US included.

etc etc etc

> But disproportionately propagating far-left content is cool.

Twitter was already caught hard coding rules to downrank content that reported Russia's invasion of Ukraine, and Russia has been continuously conducting a destabilization campaign in Europe by supporting radical fringe political movements, both far left and far right.

Stick to the facts.

Apparently while accusations of "foreign interference" are part of the background of the investigation, the actual charges that are the main current focus of the criminal investigation are (roughly translated) modification of an automated data processing system by an organized gang, and fraudulent extraction of data from an automated data processing system by an organized gang.

https://www.tribunal-de-paris.justice.fr/sites/default/files...

I'm not overly familiar with the law and I'm not defending that (I despise both the government and Elon Musk) but there are indeed laws regulating speeches and on top of that laws are flexible and the politician's power over them is quite strong. They had no problem forbidding Russian tv channels in a few days back when the war in Ukraine started

Well, in a way yes. During an election day and the day before, you are not allowed to push official communication (ie a political party can tell you "Go vote" but not "Go vote for us because we will put a stop to immigration") be it via a website, email, mail, etc. See https://presidentielle2022.conseil-constitutionnel.fr/l-elec.... So a sponsored post from a political party appearing in your timeline would be an issue due to the platform if it ignores these rules.