France Planning To Allow Use of Algorithms To Detect Extremism Online (theguardian.com) 60
Hmmmmmm shares a report from The Guardian: The French government is planning to harden counter-terrorism laws, permitting the use of algorithms to detect online extremist activity, amid a growing political row over security in the run up to next year's presidential race. The interior minister, Gerald Darmanin, said attackers were now "isolated individuals, increasingly younger, unknown to intelligence services, and often without any links to established Islamist groups." This was a growing problem for France because they self-radicalized very quickly, within days or weeks. These attackers no longer used text messages or mobile phones to communicate but instead went online or used social media direct messaging, he said. Darmanin said algorithms would allow the state to potentially pick up if a person was repeatedly searching online for a topic such as beheadings. He argued that Google and other online commercial sites already used algorithms and the state should be able to as well, with independent oversight -- despite concern from some rights lawyers that there would not be enough transparency.
"The last nine attacks on French soil were committed by individuals who were unknown to the security services, who were not on a watchlist and were not suspected of being radicalised," Darmanin told France Inter radio. This meant new methods were needed, he said, adding that of 35 attacks prevented by the state since 2017, two were stopped by intelligence work online. Since 2017, French security agencies have been able to use algorithms to monitor messaging apps. The new bill would make that experimental use permanent and extend the use of algorithms to websites and web searches. The legislation makes permanent several temporary measures in use since France's state of emergency after the Islamist terrorist attacks in 2015. It would give security agencies more power to watch over and limit the movements of high-risk individuals after release from jail, for two years rather than one.
"The last nine attacks on French soil were committed by individuals who were unknown to the security services, who were not on a watchlist and were not suspected of being radicalised," Darmanin told France Inter radio. This meant new methods were needed, he said, adding that of 35 attacks prevented by the state since 2017, two were stopped by intelligence work online. Since 2017, French security agencies have been able to use algorithms to monitor messaging apps. The new bill would make that experimental use permanent and extend the use of algorithms to websites and web searches. The legislation makes permanent several temporary measures in use since France's state of emergency after the Islamist terrorist attacks in 2015. It would give security agencies more power to watch over and limit the movements of high-risk individuals after release from jail, for two years rather than one.
Bullshit (Score:3, Insightful)
This "automated illegal conten removal" is just not working, and will be abused anyway very soon.
Re: Bullshit (Score:3, Interesting)
Re: (Score:2)
In another universe, using algorithms to detect and prosecute whatever the government doesn't like IS extremism.
But in this universe, we just call that "French".
Re: (Score:2)
What they are doing is creating digital fishing algorithm, you know, the computer says so but in reality the algorithm are tilted for and against so they can run questionable warrant fishing expeditions into people's lives, especially all political activist, those who resist propaganda, claim all sorts of nonsense as pre-crimes and then attack them through the corporate cabal sans-prosecution, unless of course prosecution can be achieved and destroy them in prison instead.
They will adjust the algorithms to
Re: (Score:1)
I seem to remember another European regime using the technology to catalog and classify its citizens according to rigid rules.
What was it called? Meh, I can't remember. I guess it's true what my friends say "Nazzy, you have a bad memory."
Re: (Score:2)
With any luck it will target anti vaxxers and anti lockdown protesters and eliminate them. Then we might have 3 deaths per million population from Covid like China instead of 1889 per million. In addition the accelerationist propaganda from left wing, right wing, religious and foreign fanatics will be deleted. Good.
Re: Bullshit (Score:5, Insightful)
Its main purpose will be to detect and remove content not deemed "appropriate" by those in power.
Note that you have made an unfounded claim based on limited information to come to a fully formed conclusion that lacks any evidence to support it. This is what is called a conspiracy theory.
Could this be abused? Yes. Is there any evidence suggesting that it will be abused? Well, you didn't present anything even suggesting they may.
Re: (Score:2)
The problem France is facing nowadays is "lone wolves", people who become extremists on their own, in their own solitude, and one day spring into action. Nobody sees them coming. They fly under the radar because they actually don't talk about their ideas, and specially not online. Good luck catching someone who doesn't post anything online with an algorithm.
The only way to catch them would be for the authorities to be informed of a change in behavior by the few friends/family they have. Usually, after the a
Re: (Score:2)
Its main purpose will be to detect and remove content not deemed "appropriate" by those in power.
Note that you have made an unfounded claim based on limited information to come to a fully formed conclusion that lacks any evidence to support it. This is what is called a conspiracy theory.
Could this be abused? Yes. Is there any evidence suggesting that it will be abused? Well, you didn't present anything even suggesting they may.
Your post is your opinion, which is no more nor less than what the OP's is. So, you too are a conspiracy theorist. See how easy that is.
Re: (Score:2)
Your post is your opinion
Literally nothing I wrote in that post was an opinion. Perhaps you need a remedial lesson in differentiating facts from opinions.
Re: Bullshit (Score:1)
Idiot. Tell us the effective difference in the end result.
Re: (Score:3)
This "automated illegal conten removal" is just not working, and will be abused anyway very soon.
Thats because the US government has allowed the DCMA to be used as a cudgel by private cartels against anyone they perceive as an enemy.
That aside, automatic content removal isn't what is being proposed here, what is being proposed is using algorithms to spot potential extremists, basically automating a few basic tasks. System throws up an alert, forwards it to a human for investigations, at which point it ends up being "Zut Alors, zis person may be a terroreest, qui should keep an eye on zem" or "eet nothi
Re: (Score:2)
IOW, "Minority Report".
Punishment before a crime is ever committed, determined by an "algorithm" which codifies the bias of its creators. In China it is called the "Social Credit Score". You can't even get toilet paper at a public restroom of the facial ID algorithm identifies you as a person with a low score. Imagine walking home with a dirty butt, unable to even ride a bus or hire a taxi. Those with low scores soon learn to carry their own roll of TP and a canteen of water.
Re: (Score:3)
Re: (Score:2)
I note that my YouTube comments account has been terminated twice for calling morons - morons. Apparently it is hate speech to question the sanity of people. You never get a reason for being terminated either so I actually have no idea what I was terminated for. Screw them, new accounts are easy to create. I am also noticeably nastier after each ban with the things I say. Why hold back when you have no idea what you are being banned for.
Re: (Score:2)
And if they drive the extremist talk underground then they won't be able to keep an eye on it.
I expect gov't 3-letter type agencies to monitor the web for extremism and to watch closely any potential terrorists. Telling them they are listening in to conversations is a dumb idea.
Chasing the fifth column (Score:4, Informative)
This situation and the fight against it are increasingly reminiscent of the fifth column [wikipedia.org] rhetoric, which is all the more disturbing because the term was coined during a civil war and widely used in the wars that followed.
It working! (Score:2)
Look - slashdot 0 comments!
Efficiency (Score:5, Funny)
"Computer! Plese give me a list of all extremist pages on the web!"
*Found 5.27 billion entries."
"Hmm, guess I'll have to narrow it down a bit: Religious. Violent. Self-Righteous. Authoritarian. Obscurantist. Here we go!"
*Found 5.27 billion entries."
"Screw it, I'll just graze reddit."
Extremism Isn't Online. (Score:2)
It is inside someone's head until it becomes reality.
Re: (Score:2)
Maybe we
Re: (Score:2)
get them some help
Puberty blockers?
Re: (Score:2)
I note that there are sophisticated propaganda outlets online which pump out hate speech day in and day out against their chosen enemies. Some are Islamic State, some enemy States and some are just political parties that rely on hatred of the opposition for votes. People seeking an identity are easily drawn into this mess and end up with terror cult mindsets. These outlets are not Catcher in the Rye, they are there to push the hate button daily - usually for political influence but easily the cause of mass
Re: (Score:2)
It is inside someone's head until it becomes reality.
Exactly. School shooters for instance don't just appear overnight. It starts with a bunch of lonely people and a small percentage of those become depressed and a small percentage of those become suicidal and a small percentage of those decide to take other people with them. Or it might start with someone being bullied and a small percentage of those go down the rabbit hole too. But the rabbit hole is still a series of funnels with literally millions of people at the top. It's kindof like radiation ex
Current algo did not catch the latest terrorists (Score:5, Interesting)
Basically, the current algorithm detects if you call Syria on your phone.
It did not detect if you use FB Messenger (etc.) to talk to suspect persons/locations.
Currently it does NOT detect if you spend your day on the net looking at beheading videos from ISIS.
The proposed algorithm would flag suspect things, then the police will have to contact a judge and a prosecutor and decide if the data should be de-anonymous.
In my opinion, GAFA already has the data, so why not the anti terrorist organisation too?
Re: Current algo did not catch the latest terroris (Score:4, Interesting)
Oh kids. In the early 2000s, at least in Germany, watching shocker videos, including 3guys1hammer and beheadings, were a national coming of age ritual for young people, especially boys.
Basically a test to how strong your mind was and what it could handle, to train you for being able to keeo yourself together in real life, even in the case shit goes down. (You know... exactly what the special snowflakes did not learn thanks to their safe spaces.)
And it was a test exactly *because* one did not like it.
It's not new though. In the decades before, watching splatter movies and generally all banned movies was the equivalent.
Not to campare the teo, but just for the sake of an analogy, it's like the difference between the recluse perv who never leaves his basement and watches child porn, and the one who actually *makes* the child porn.
The former requires immediate intense therapy, of course. But it's the latter who is at the top of the list of who one needs to catch.
The former might not even want to have these feelings, let alone harm children. The latter does.
Re: (Score:2)
Oh kids. In the early 2000s, at least in Germany, watching shocker videos, including 3guys1hammer and beheadings, were a national coming of age ritual for young people, especially boys.
So an entire generation intentionally suffered self-induced PTSD?
That explains a few things about Germany...
Who is Darmanin (Score:4, Informative)
In English, but a lot less detailed [wikipedia.org]
In short : a racist homophobic rapist who became Interior Minister (mostly cops minister)
Re: Who is Darmanin (Score:2)
So France's Schäuble. (Germany's Cheney.)
Re: (Score:2)
Re: (Score:2)
This is like 'Yo president' jokes in reverse, everyone needs to hop in on this.
Re: (Score:2)
Although I really dislike the French president, I don't think bashing him on his choice of an older wife is okay. I mean, he is one of the few people in power who does not seem to use his power for abusing women. Previous French presidents (except maybe Sarkozy and De Gaulle) all had what would be considered very problematic behavior in today's society. And let's not talk about Darmanin.
Although nobody should receive praise for not being an abuser, I thought it might be useful to mention that.
spontaneous fission (Score:1)
Remember the Bastille,
Algorithm Search here: (Score:5, Funny)
Re: Algorithm Search here: (Score:2)
Would be funnier, if you had put a big "NOT" in front.
(We established that an algorithm or neural net cannot detect sarcasm or many other forms of negation unless it lived an entire human life in the same society and community as the speaker.)
Re: (Score:2)
or "... in Minecraft!"
Oh great. Institutionalized prejudice. (Score:2)
Bureaucratic one-rule-fits-all prejudice too.
With no human to intervene in cases where common sense dictates exceptions or bugs happen.
Because ...
Spongebob_magic_rainbow_meme.gif
*algorithms*!
Last but not least:
Picard: Computer! Define "extremism"!
Computer: Everthing that the established dicators at the time deemed too far outside the goosestepping line. It was a convenient tool, to put everyone who does not obey in with terrorists and criminals, and give the citizens a convenient scapegoat for all problems,
Computer says ... (Score:2)
... that you are an extremist. Therefore no human can be blamed for the decisions that follow.
Same with the anti-cheating software that universities used during the lock-down for exams taken from home: Computer says you cheated. You cannot sue the university, go ahead and sue the computer.
10 days ago: EU proposes restrictions on AI (Score:1)
What is extremism? (Score:5, Insightful)
It's a genuine question I have and I think this is really what a lot of countries are going to have to seriously address.
There are those who think extremism is merely the use of violence. You can hold whatever view you want, as long as you don't act violently. On some surface level, that sounds interesting. The problem comes when it actually hit human nature. People want to follow their beliefs. A vegan doesn't want to see animals slaughtered. You create a very tense and conflict prone society as people's core values are challenged. I only use vegans here as an example to hopefully avoid hotter groups. But who knows maybe Vegans are just as hot. If you get a group of vegans who genuinely belief animal slaughter is murder... do you not think they'd think themselves righteous if they took stronger, possible violent measures to achieve that goal.
Now, I'll venture a bit more. Take abortion as it relates typically to religious group. The claim is often that abortion is murder. Interesting view. But ponder that for a second. If you really believed abortion is murder, would you not be righteous is trying to stop it? If the government was lining up 5 year olds in government schools and shooting them... who of us would not think we should possibly violently revolt against that?
I speak from personal experience here. I grew up very religious (conservative Sunni Muslim). To put it lightly I genuinely felt strong moral and righteous motivations in certain ways. I do not joke when I say, a wrong turn here or there or the wrong associations here or there, and who knows what path my life would have led. As a teengager, I genuinely had vision of being a holy warrior and liberating Palestine. I've never been to Palestine. I only grew up in Africa and then in Canada. I genuinely thought the fiat monetary system of the West was evil and that only a return to the Islamic gold standard or something else could liberate us from the evil of big finance.
That same drive that perhaps makes people call oil companies evil or big banks evil or donald trump evil or socialism evil or communism evil or homosexuality evil or the patriarchy is evil or feminism is evil... is what drives so many of us. Whatever cause we think is righteous. Someone might fight feminism as they see it as a threat to the righteousness of family and god life. Others might fight religion as they see it as a threat to their freedoms.
The way I've come to see it is that violence is not what we need to worry about. Everyone and every group is willing to use violence to correct what they see as righting a wrong. How much would be blame the Uyghurs in China if they used violence to revolt against China? China would naturally use violence in return or state control violence to keep China together. The US civil war is another classic case. Also never forget the state is always in a state of violence against those who oppose it. Again from personal experience. You might think of Canada as this peaceful place, but again, back in the day I viewed the Canadian state (government, police, banks, military...) as oppressing We were living under Western subjugation preventing us from living a real Muslim life under Islamic rules. It's just incorrect to think that just because the state isn't actively committing genocide that it's not using force. The threat of force is just that... force.
We often try to reduce beliefs to trivial things... until they matter. But they do matter. Whatever you feel so strongly about... chances are your worst enemy feels strongly and righteously about their beliefs.
In my view, we either try and actually reduce beliefs in society so people don't care as much. Because any strong belief in and of itself is 'extremism'. Or you minimize the number of beliefs in your society. In a sense, you only allow one extremist thought in your society.
You either say that no one can really have strong beliefs. You're only allowed to be a casual Muslim, Christian, Feminist, Environmentalist, Socialist... You can be Muslim. Follow
Re: (Score:2)
I literally just gave 2 solutions in my post.
1. prevent people from having serious beliefs. Some countries like say Singapore do this a fair bit. Publications of controversy around religion is largely banned. Free speech is limit in any way that can prevent harmony. They actively manage trying to ensure social harmony.
2. Have only one supremacy. This one is easy as so many countries do it. But to pick an easy example. Saudi Arabia is an Arab Islamic Supremacy. They make no bones about it. You know what you'
Context, False positives, and bad algorithms (Score:2)
How does an algorithm determine context between someone ...
* discussing/searching history,
* using humor,
* actual planning of violence, and
* trolling?
Any sort of "automated" system is always going to have false positives due to lack of context.
The excuse is always "Think of the children" / safety but this is a dangerous slippery slope where it seem innocuous at first and slowly expands. Who keeps the system in check? Is the source code available so people can
Not too worried. (Score:2)
Re: (Score:2)
Yes but we spend our taxes on those systems, which is bad. I'd rather pay my taxes to hire more police, or (maybe more productive in the long term) better teachers.
Extremism (Score:2)
"Extremism in defense of liberty is no vice." -- Barry Goldwater
"Liberte, egalite, fraternite." -- French Revolution.
"Vive France Libre" -- Charles DeGaulle.
"Au revoir, la France." -- Macron.
the French government is the extremists (Score:2)
It's too expensive and they're terrible at it.
[alt:"music notes emoji"] (Score:1)
Cue up Black Sabbath - Paranoid.