X Raided in France

ThatRobGuy · Sunday at 9:49 AM

Stopped_lurking said:
Perhaps, or they think they can make a case stick to X, but they don't think they can make it stick against Mistral.

Seems like Mistral offering up suggestions to users on how to groom kids online for a real-world meetup and teaching people how to make homemade explosives and chemical weapons would suggest a serious lack of guardrails. (and being 60 times more likely to do so than the other major providers like OpenAI and Anthropic should be a red flag). If we were talking about a credit card software or medical software provider, and they were 60 times more likely to expose customers CC numbers or SSN number/medical history, they'd be all over it)

I think sometimes the obvious answer is the correct one... Going after the Euro AI companies would be directly counter productive to what Macron's stated economic goals were. Which was to foster the growth and competitiveness of European AI companies.

Stopped_lurking said:
Were their servers in France? Do they have the data still? Do french police and prosecutors believe they can make a case against them?

Yes, they operate out of 4 different data centers within a 30 minute vicinity of their main HQ (one of which being almost exclusively dedicated to their stuff, the others are shared data centers)

Stopped_lurking said:
I was thinking about the American car manufacturer example. Or any example that we can look at in hindsight and determine what actually happened.

It was just a hypothetical example...

A more real-world example would be the scrutiny and enforcement actions against TikTok, for things that American social media companies were doing.

A 2022 Consumer Reports study found that TikTok uses the same data-tracking practices as Facebook/Meta and others, collecting information about online and offline activities, location, what websites you're visiting, and what links you click on.

Both The Washington Post and The University of Toronto's Citizen Lab examined TikTok independently and concluded that TikTok does not appear to collect any more data than typical mainstream social networks—in fact, Facebook and Google both collected more personal data from users than TikTok does.

ThatRobGuy · Sunday at 9:52 AM

Hans Blaster said:
Never heard of any of them. I see no reasons to know the names of these companies that will all fail soon.

With a few them having well over half the market share in their areas of specialty, it's doubtful they fail (at least in the European markets)

Especially when the leadership in France is willing to leverage selective enforcement to run their competitors out of town.

If Trump started sending in the goon squads to raid Toyota and Honda facilities in the US, to the benefit of Ford and GM... Ford and GM will survive despite making sub-par vehicles compared to the Japanese auto makers.

Stopped_lurking · Sunday at 9:59 AM

ThatRobGuy said:
Seems like Mistral offering up suggestions to users on how to groom kids online for a real-world meetup and teaching people how to make homemade explosives and chemical weapons would suggest a serious lack of guardrails. (and being 60 times more likely to do so than the other major providers like OpenAI and Anthropic should be a red flag). If we were talking about a credit card software or medical software provider, and they were 60 times more likely to expose customers CC numbers or SSN number/medical history, they'd be all over it)

I think sometimes the obvious answer is the correct one... Going after the Euro AI companies would be directly counter productive to what Macron's stated economic goals were. Which was to foster the growth and competitiveness of European AI companies.

The obvious answer is, if they think the can make the charges stick they would. If they can't they won't go after them.

ThatRobGuy said:
Yes, they operate out of 4 different data centers within a 30 minute vicinity of their main HQ (one of which being almost exclusively dedicated to their stuff, the others are shared data centers)

Stability AI is not headquartered in France. How many of their servers are in France? Do you believe their servers contain any illegal content?

ThatRobGuy said:
It was just a hypothetical example...

A more real-world example would be the scrutiny and enforcement actions against TikTok, for things that American social media companies were doing.

A 2022 Consumer Reports study found that TikTok uses the same data-tracking practices as Facebook/Meta and others, collecting information about online and offline activities, location, what websites you're visiting, and what links you click on.

Both The Washington Post and The University of Toronto's Citizen Lab examined TikTok independently and concluded that TikTok does not appear to collect any more data than typical mainstream social networks—in fact, Facebook and Google both collected more personal data from users than TikTok does.

So because the US government does it, the french government is doing it. Isn't that just straight projection?

ThatRobGuy · Sunday at 10:48 AM

Stopped_lurking said:
The obvious answer is, if they think the can make the charges stick they would. If they can't they won't go after them.

Or, they won't go after them and will let it slide to a degree for domestic companies as to not impact their economic goals.

Stopped_lurking said:
Stability AI is not headquartered in France. How many of their servers are in France? Do you believe their servers contain any illegal content?

That I don't know (in terms of which data centers they operate out of), it's also quite possible they use a cloud provider in which case it would likely be a set up where they have instances stood up in EU-London, EU-Paris, and EU-Dublin for redundancy purposes.

Stopped_lurking said:
So because the US government does it, the french government is doing it. Isn't that just straight projection?

I think it's an acknowledgement of reality. You don't think every country has a vested interest in trying to make sure their companies do better than companies from other countries?

Macron has not been coy about his vision of "Europe First" for AI

The 15 days that upended Macron’s vision for European AI

The French tech world is reeling from Donald Trump’s Stargate announcement and DeepSeek’s recent breakthrough.

www.politico.eu

When the United States says "America First", it conveys a certain notion does it not? (perhaps a blend of patriotism/nationalism?)

Macron has expressed a "Europe First" vision for AI.

The president’s message was one of “European preference,” said a French government official. “We are not going to control what companies do, but European patriotism is at stake.”

In the eyes of the leader who prophesied the death of Europe, projects like the data center are crucial, as European civilization and its values that are at stake. If Europe cannot harness the power of AI, clean tech and other new technologies, Macron has argued, its languages and democratic traditions will be imperiled.

When Macron starts saying things that are basically to the effect of "we've been getting pushed around by other countries for too long, it's time to push back", and says "patriotism is at stake", and is equating not getting leg up in this AI race with "the death of Europe" and "the demise of our values, traditions, and language"

Essentially, anchoring the sentiments of national identity, patriotism, and tradition to his tunnel-visioned goal in the public discourse...

Whose rhetoric does that resemble just a little bit? It borders on being one step away from him popping on a baseball cap that says "Make France Great Again"

Stopped_lurking · Sunday at 11:09 AM

ThatRobGuy said:
Or, they won't go after them and will let it slide to a degree for domestic companies as to not impact their economic goals.

Do you have any reason to believe that this is the case?

ThatRobGuy said:
That I don't know (in terms of which data centers they operate out of), it's also quite possible they use a cloud provider in which case it would likely be a set up where they have instances stood up in EU-London, EU-Paris, and EU-Dublin for redundancy purposes.

So why would you expect the french police to go after them?

ThatRobGuy said:
I think it's an acknowledgement of reality. You don't think every country has a vested interest in trying to make sure their companies do better than companies from other countries?

Not by looking the other way in CSAM cases, no.

ThatRobGuy said:
Macron has not been coy about his vision of "Europe First" for AI

Where did he say "Europe first"? It is a subheading in the article, not anything he is quoted saying.

ThatRobGuy said:
The 15 days that upended Macron’s vision for European AI

The French tech world is reeling from Donald Trump’s Stargate announcement and DeepSeek’s recent breakthrough.

www.politico.eu

When the United States says "America First", it conveys a certain notion does it not? (perhaps a blend of patriotism/nationalism?)

Macron has expressed a "Europe First" vision for AI.

The president’s message was one of “European preference,” said a French government official. “We are not going to control what companies do, but European patriotism is at stake.”

In the eyes of the leader who prophesied the death of Europe, projects like the data center are crucial, as European civilization and its values that are at stake. If Europe cannot harness the power of AI, clean tech and other new technologies, Macron has argued, its languages and democratic traditions will be imperiled.

When Macron starts saying things that are basically to the effect of "we've been getting pushed around by other countries for too long, it's time to push back",

Where did he say that? It is not in the articles you quoted.

ThatRobGuy said:
and says "patriotism is at stake"
, and is equating not getting leg up in this AI race with "the death of Europe" and "the demise of our values, traditions, and language"

Where did he say anything about the "the demise of our values, traditions, and language"? This is not in the articles you linked.

Are you using quotation marks even if you are not actually quoting someone?

ThatRobGuy said:
Essentially, anchoring the sentiments of national identity, patriotism, and tradition to his tunnel-visioned goal in the public discourse...

Whose rhetoric does that resemble just a little bit? It borders on being one step away from him popping on a baseball cap that says "Make France Great Again"

Hans Blaster · Sunday at 12:16 PM

ThatRobGuy said:
With a few them having well over half the market share in their areas of specialty, it's doubtful they fail (at least in the European markets)

The "market" for this slop is very thin.

ThatRobGuy said:
Especially when the leadership in France is willing to leverage selective enforcement to run their competitors out of town.

I have no idea what level of evidence is available to get a warrant in France and neither does any one on this thread.

ThatRobGuy said:
If Trump started sending in the goon squads to raid Toyota and Honda facilities in the US, to the benefit of Ford and GM... Ford and GM will survive despite making sub-par vehicles compared to the Japanese auto makers.

Don't give him any ideas.

loveofourlord · Sunday at 2:33 PM

Oompa Loompa said:
I agree. It is repulsive. I think any pornography is repulsive. But just to play devil's advocate, if some sick pedophile uses AI to generate child porn, who is the victim? What is the crime? I mean, which would you prefer, a pedophile abducting a child to force them into a psychopathic photo shoot that would leave them traumatized for the rest of their lives, or some guy getting an AI to give them what they need without the trauma? This is an ethical dilemma for many. My opinion is that lust is never satisfied. So AI child porn would just be a temporary solution before these sick individuals will not even be satisfied with that and seek human victims anyway.

this isn't simply child porn which be an issue in itself, this is taking photos of REAL kids, then using grok to remove the clothes. THere are literal victims here.

loveofourlord · Sunday at 2:44 PM

Something to point out, grok wouldn't be able to make child porn, unless it was trained on child porn. AI only can make what it's already given.

ThatRobGuy · Monday at 7:11 AM

Stopped_lurking said:
Do you have any reason to believe that this is the case?

So why would you expect the french police to go after them?

Not by looking the other way in CSAM cases, no.

Where did he say "Europe first"? It is a subheading in the article, not anything he is quoted saying.

Where did he say that? It is not in the articles you quoted.

Where did he say anything about the "the demise of our values, traditions, and language"? This is not in the articles you linked.

Are you using quotation marks even if you are not actually quoting someone?

His quotes actually were referenced in the Politico piece linked, but here are some others

EU must avoid becoming tech 'vassal' of US and China, Macron says

French President Emmanuel Macron warned Tuesday that Europe must avoid becoming a tech "vassal" of the US and China, urging a "European preference" as German Chancellor Friedrich Merz echoed calls in…

www.france24.com

France leads push for widespread adoption of European AI

“We have to stop being naïve.” That was French President Emmanuel Macron’s message as he called on public authorities and industry to support European AI systems, during the Adopt AI Summit in Paris on November 25.The event, billed by organisers as the Davos for AI, in reference to the World...

sciencebusiness.net

https://www.bloomberg.com/news/articles/2021-09-29/is-macron-straining-relations-with-europe-as-he-seeks-strategic-autonomy

He's overtly called for showing preferential treatment for European AI companies

And his "Digital sovereignty" agenda has been in a thing for the past few years

France dumps Zoom and Teams as Europe seeks digital autonomy from the U.S.

European governments are moving away from U.S. tech giants, opting for domestic or open-source alternatives.

www.latimes.com

The French government made this announcement:
...announced last week that 2.5 million civil servants would stop using videoconferencing tools from U.S. providers — including Zoom, Microsoft Teams, Webex and GoTo Meeting — by 2027 and switch to Visio, a homegrown service.

The objective is “to put an end to the use of non-European solutions, to guarantee the security and confidentiality of public electronic communications by relying on a powerful and sovereign tool,” the announcement said.

Stopped_lurking · Monday at 7:47 AM

ThatRobGuy said:
His quotes actually were referenced in the Politico piece linked, but here are some others

There is two Politico articles linked, in which one do you find the following sentence?

"the demise of our values, traditions, and language"

Also, I couldn't find the word pushed in either of them.

ThatRobGuy said:
EU must avoid becoming tech 'vassal' of US and China, Macron says

French President Emmanuel Macron warned Tuesday that Europe must avoid becoming a tech "vassal" of the US and China, urging a "European preference" as German Chancellor Friedrich Merz echoed calls in…

www.france24.com

France leads push for widespread adoption of European AI

“We have to stop being naïve.” That was French President Emmanuel Macron’s message as he called on public authorities and industry to support European AI systems, during the Adopt AI Summit in Paris on November 25.The event, billed by organisers as the Davos for AI, in reference to the World...

sciencebusiness.net

https://www.bloomberg.com/news/articles/2021-09-29/is-macron-straining-relations-with-europe-as-he-seeks-strategic-autonomy

He's overtly called for showing preferential treatment for European AI companies

And his "Digital sovereignty" agenda has been in a thing for the past few years

France dumps Zoom and Teams as Europe seeks digital autonomy from the U.S.

European governments are moving away from U.S. tech giants, opting for domestic or open-source alternatives.

www.latimes.com

The French government made this announcement:
...announced last week that 2.5 million civil servants would stop using videoconferencing tools from U.S. providers — including Zoom, Microsoft Teams, Webex and GoTo Meeting — by 2027 and switch to Visio, a homegrown service.

The objective is “to put an end to the use of non-European solutions, to guarantee the security and confidentiality of public electronic communications by relying on a powerful and sovereign tool,” the announcement said.

Neither of which means that the French police and prosecutors are involved looking the other way in regards to CSAM.

ThatRobGuy · Monday at 8:16 AM

Stopped_lurking said:
There is two Politico articles linked, in which one do you find the following sentence?

"the demise of our values, traditions, and language"

Also, I couldn't find the word pushed in either of them.

In this paragraph of the Politico article...

As well as this one:

And if you noticed how I worded my post when referencing that:
things that are basically to the effect of "we've been getting pushed around by other countries for too long, it's time to push back", and says "patriotism is at stake", and is equating not getting leg up in this AI race with "the death of Europe" and "the demise of our values, traditions, and language"

Stopped_lurking said:
Neither of which means that the French police and prosecutors are involved looking the other way in regards to CSAM.

No, but that combined with the fact that Mistral and StabilityAI have had these same issues (and to a larger scale)

And the response was to raid Grok for it...

Meanwhile, Macron is hosting swanky events for Mistral CEO Arthur Mensh...

And ironically enough, Mensch's stance on AI dovetails into the very thing that people are debating and viewing as problematic about Musk's stance

Per Medium:
The controversy primarily revolves around the lack of moderation mechanisms in the model. Unlike other AI models that have safeguards to prevent the generation of harmful or inappropriate content, Mistral’s LLM can generate responses to any query without moderation. Independent tests revealed that the model could provide detailed instructions on harmful or illegal activities, such as creating a bomb, which other models from Meta, OpenAI, and Google refuse to answer. This has raised significant safety concerns and ignited a heated discussion on the responsibility of AI developers to ensure the safety of their models.

Mistral AI’s co-founder, Arthur Mensch, has argued that it is the responsibility of developers and users, not the builders of the models, to ensure that AI applications are safe. He views their models as tools, similar to programming languages, which can be used for both beneficial and harmful purposes. However, this stance has been met with criticism.

So, to recap:
X/Grok Musk: has the philosophy of "it's a tool, it's on the user to use it ethically, we don't want to be in the moderation business" (and people can and have generated unethical things with it as a result)

France: "We're raiding your offices"

Mistral Mensch: has the philosophy of "it's a tool, it's on the user to use it ethically, we don't want to be in the moderation business" (and people can and have generated unethical things with it as a result)

France: "We'd like to invite you to be a keynote speaker at the $500/plate AI Action Summit gala...can you pose with the president for a picture?".

Stopped_lurking · Monday at 8:47 AM

ThatRobGuy said:
In this paragraph of the Politico article...

View attachment 376339

As well as this one:
View attachment 376340

And if you noticed how I worded my post when referencing that:
things that are basically to the effect of "we've been getting pushed around by other countries for too long, it's time to push back", and says "patriotism is at stake", and is equating not getting leg up in this AI race with "the death of Europe" and "the demise of our values, traditions, and language"

This last one was not in the text. The way you worded it I would only expect the first one to be your paraphrase.

Why paraphrase at all, if you can quote instead? Now, you are injecting your interpretation of the text instead of what was actually said.

ThatRobGuy said:
No, but that combined with the fact that Mistral and StabilityAI have had these same issues (and to a larger scale)

What connection is there between french police and stability AI?

ThatRobGuy said:
And the response was to raid Grok for it...

Because they think they can make the charges stick, is my guess.

Grok might be even easier to manipulate for generating CSAM. The user base for Grok is larger, perhaps they had reason to believe they can actually find the generated pictures on their servers.

ThatRobGuy said:
Meanwhile, Macron is hosting swanky events for Mistral CEO Arthur Mensh...
View attachment 376341

And ironically enough, Mensch's stance on AI dovetails into the very thing that people are debating and viewing as problematic about Musk's stance

Per Medium:
The controversy primarily revolves around the lack of moderation mechanisms in the model. Unlike other AI models that have safeguards to prevent the generation of harmful or inappropriate content, Mistral’s LLM can generate responses to any query without moderation. Independent tests revealed that the model could provide detailed instructions on harmful or illegal activities, such as creating a bomb, which other models from Meta, OpenAI, and Google refuse to answer. This has raised significant safety concerns and ignited a heated discussion on the responsibility of AI developers to ensure the safety of their models.

Mistral AI’s co-founder, Arthur Mensch, has argued that it is the responsibility of developers and users, not the builders of the models, to ensure that AI applications are safe. He views their models as tools, similar to programming languages, which can be used for both beneficial and harmful purposes. However, this stance has been met with criticism.

So, to recap:
X/Grok Musk: has the philosophy of "it's a tool, it's on the user to use it ethically, we don't want to be in the moderation business" (and people can and have generated unethical things with it as a result)

France: "We're raiding your offices"

Mistral Mensch: has the philosophy of "it's a tool, it's on the user to use it ethically, we don't want to be in the moderation business" (and people can and have generated unethical things with it as a result)

France: "We'd like to invite you to be a keynote speaker at the $500/plate AI Action Summit gala...can you pose with the president for a picture?".

ThatRobGuy · Monday at 9:34 AM

Stopped_lurking said:
Because they think they can make the charges stick, is my guess.

Grok might be even easier to manipulate for generating CSAM. The user base for Grok is larger, perhaps they had reason to believe they can actually find the generated pictures on their servers.

I'm asserting that it's less about "can they make charges stick", and exists more in the realm of "do they even have a desire to bring charges in the first place for extraneous reasons"

A noted, Mistral's issues (stemming from their weak guardrails and the overall philosophy of their CEO regarding regulation and moderation) have been established since 2024.

Macron's initiatives and vision for EU being a major player in the AI realm and promoting "European preference" has a lot riding on Mistral and StabilityAI succeeding.

As the idiom goes, you don't shoot the horse you've hitched your wagon to.

Stopped_lurking · Monday at 12:54 PM

ThatRobGuy said:
I'm asserting that it's less about "can they make charges stick", and exists more in the realm of "do they even have a desire to bring charges in the first place for extraneous reasons"

This seems very conspiratorial. I don't think the prosecutors office would let those things slide if they thought they could make it stick.

ThatRobGuy said:
A noted, Mistral's issues (stemming from their weak guardrails and the overall philosophy of their CEO regarding regulation and moderation) have been established since 2024.

Macron's initiatives and vision for EU being a major player in the AI realm and promoting "European preference" has a lot riding on Mistral and StabilityAI succeeding.

As the idiom goes, you don't shoot the horse you've hitched your wagon to.

ThatRobGuy · Monday at 1:54 PM

Stopped_lurking said:
This seems very conspiratorial. I don't think the prosecutors office would let those things slide if they thought they could make it stick.

Conspiracy would imply that it's something that's being kept secret.

When the president of France has openly stated they're seeking "Euro tech/AI Sovereignty" and "European preference" over American companies...

And there are two companies (one American and one French) that both have similar issues...and both with CEOs that have expressed a hands-off philosophy with regards moderation and both asserted "it's a tool, it's on the users what they decide to do it".

And he takes one to court, and takes the other one out to dinner...

It doesn't exactly take Columbo to solve that mystery.

Contrary to popular belief, Americans aren't the only ones that are susceptible to tunnel vision with regards to nationalism and global supremacy-based economic goals.

Stopped_lurking · Monday at 10:34 PM

ThatRobGuy said:
Conspiracy would imply that it's something that's being kept secret.

The conspiracy would be that there is any unknown influence or secret instructions to police and prosecutors to not start a case against a french AI startup even if they thought they could get a conviction.

ThatRobGuy said:
When the president of France has openly stated they're seeking "Euro tech/AI Sovereignty" and "European preference" over American companies...

And there are two companies (one American and one French) that both have similar issues...and both with CEOs that have expressed a hands-off philosophy with regards moderation and both asserted "it's a tool, it's on the users what they decide to do it".

And he takes one to court, and takes the other one out to dinner...

It doesn't exactly take Columbo to solve that mystery.

Neither of that is evidence of that the french police and prosecutors are looking the other way.

ThatRobGuy said:
Contrary to popular belief, Americans aren't the only ones that are susceptible to tunnel vision with regards to nationalism and global supremacy-based economic goals.

We hope the site problems here are now solved, however, if you still have any issues, please start a ticket in Contact Us

X Raided in France

Part of the IT crowd

Part of the IT crowd

Well-Known Member

Part of the IT crowd

Well-Known Member

Call Me Al

Newbie

Newbie

Part of the IT crowd

Well-Known Member

Part of the IT crowd

Well-Known Member

Part of the IT crowd

Well-Known Member

Part of the IT crowd

Well-Known Member

Similar threads