Hacker News new | past | comments | ask | show | jobs | submit login
Disclosing networks of state-linked information operations we’ve removed (blog.twitter.com)
179 points by mpweiher 8 months ago | hide | past | favorite | 81 comments

I wonder if Twitter also does the same for Western governments ops or only for “bad guys” like Russia, China and Turkey.

In France, during the “Yellow Vests” movement, there's has been a lot of progovernment activity on Twitter from accounts which were obvious fake ones: with a photo stock profile picture and a username made of “(often vintage) French first name+ 5 or 6 random figures”.

I suspect this is frequent among western countries too but I don't have evidence except for my country, if you have such stories for your country, please share.

There are quite a few people and organisations analysing social media accounts. They have, on occasion, exposed botnets, or networks of fake accounts that are manually run.

I know of a few here in Germany, and I have no doubt that they would gleefully publish any US government networks they come across. And with the extreme polarisation in the US, there should be plenty of people or publications on either side to do the same.

The fact that, as far as I can tell, these tend not to be bots but manually-run fake accounts might point to an explanation of the relative lack of such activity: it might just be too expensive. And when it comes to networks targeting other countries, every country that isn't natively English- or Spanish-speaking is relatively save because you aren't going to find hundreds of Americans speaking Russian willing to work for low wages and reliably keep quiet about it.

That, of course, is in addition to the obvious reason that democratic countries by their very nature just don't do stuff like that. But I have a hunch the ever-cynical crowd here would take offence at the idea that some countries and governments sometimes don't behave in the worst possible way.

> That, of course, is in addition to the obvious reason that democratic countries by their very nature just don't do stuff like that

They absolutely do, with varying levels of transparency. I remember getting annoyed when the Scotland Office were posting anti-independence propaganda in the form of Buzzfeed listicles: https://www.buzzfeed.com/youdecide2014/scotland-the-uk-10-my...

Now, a couple of clicks will reveal that "you decide 2014" do say that they're the UK government, but it's not instantly obvious when you're looking at the Buzzfeed article.

The current UK Transport secretary, Grant Shapps, has previous for running a firm to game Google AdSense: https://www.theguardian.com/politics/2012/sep/02/grant-shapp...

There have been several announcements of UK web propaganda operations: https://www.realwire.com/releases/British-Army-to-present-at... / https://www.theguardian.com/uk-news/2015/jan/31/british-army...

> democratic countries by their very nature just don't do stuff like that

Democracy isn't a black and white things. Turkey is a democracy, no matter how corrupt and authoritarian Erdogan is. And democracy-wise France is arguably closer to Turkey than to the UK anyway.

Really? With UK’s oppressive anti-privacy laws? Anyway, I live in Switzerland, and we are the only real (direct) democracy out there. :)

Democracy doesn't preclude oppression or promote privacy. It also doesn't require that all citizens have the ability to vote, that all living in the country are citizens, or that citizens be free. Just like most forms of government (so far as I know, the only pure forms of communist/republican/democratic/etc government in current use are small scale town/village level, so this doesn't really mean much). I also don't have much knowledge on how most current governments work at any level of detail, so feel free to tell me if I'm wrong.

democratic countries by their very nature just don't do stuff like that

I am afraid I don't have a reference, but some years ago I read about a request for bids on persona management software on fbo.gov, so one person could run many accounts.

> democratic countries by their very nature just don't do stuff like that.

Being too cynical is certainly as wrong as being too naive. This sound a bit like that latter. I would think the Russian influence stories were a form of simple propaganda. That was not a mistake by intelligence because they already knew long before. They didn't plant bots though, that is too expensive for most countries, especially if the language barrier is that high. China might employ some people, but it is unfeasible for most countries.

To be honest, I think people accusing others to be bots are wrong in 99.9% of cases.

It's not like your need a lot of text. Do you really think that producing 10k different messages is "expensive" for an intelligence group? (Not that you need to write that many with some templates and randomisation)

Yes, lots of people heard about bots and use the accusation incorrectly. But in many cases actual sockpuppets are easy to spot. There's lots of accounts used only for propaganda for example and switch from Chinese-supporting messages when there's something happening in China to Russia supporting content following some news, to Trump supporting the next month before elections. They don't even bother deleting old messages, because most people don't check / doesn't care.

Twitter didn’t even remove ISIS content until the US government finally yielded that it wasn’t a grassroots rebellion and that it was a bad organisation.

I reported dozens of accounts spreading propaganda in 2014 and it was never actioned. This was everything from handbooks through to assassination clips.

It makes it worse when you consider that Twitter was their primary media and, had Twitter not followed the US line and actioned the genuine reports, then the impact ISIS had could have been reduced.

But that said - it’s not like Twitter is alone in their adherence to US gov foreign policy. All mainstream US media (WaPo, WSJ, NY Times, Economist et al) do the same thing. There is hardly any independence in international reporting.

Definitely, I've seen a lot of fake accounts created in this format Name+series of numbers

But apparently there are some real accounts that follow this pattern. I have no idea why. Maybe Twitter suggests this username?

But yeah, stock photo/"generic"/drawing or no photo are more telling.

I'm surprised they're not disclosing the US ones.

AFAIK twitter doesn't even ban US state operated psy-ops, not even ones targeting US citizens.

They should require all parties posting on behalf of a state or as part of a contract for a state identify themselves...

Given the revolving door at Twitter, or other mainstream media outlets for that matter, this isn't surprising.

It's easy and safe to disclose networks of states that have no power over you. The information operations that are not disclosed are the ones you really need to worry about.

Or the Mexican ones, who even tag themselves with the #RedAMLO hashtag.

Yes, the greater context is that there are many entities exhibiting this sort of behavior online, not just countries and not just those three countries.

Do you mean the ones promoting this or that cause and who are normally aligned with one or the other major parties or significant pols?

Whatabout... the US.

Wouldn't disclosing any US sponsored psy-ops be illegal by US law?

The press is protected by precedent (and the first amendment in general); they can publish any material they want, no matter how classified it is. The person who gets in trouble is the one who broke the laws and gave the material to the press.

The government has asked things not to be published in the past, particularly in the aeropspace sector. Sometime the press has complied, sometimes it hasn't.

Hell, the organization I used to work for has at least 3 pieces of TS material on their wikipedia page. Years back I brought my concerns to the security manager and they said "the worst thing you can do with classified material is try to remove it, because you're just validating that it's classified and correct."

> The press is protected by precedent

Is journalism a protected profession? In my country journalists are not, so they have the same rights/obligations as anyone.

> The person who gets in trouble is the one who broke the laws and gave the material to the press.

What if the information leaked contained evidence of crime? We have still people in prison for that...

I can't imagine US being all that significant.

Biden, Trump and various members of both camps command more than enough attention through their own Twitter accounts.

Likewise pushing political agendas is largely superfluous given you can just lobby politicians directly without needed public support.

You're not considering foreign operations. If included, I bet the US would be at the forefront.

Luckily the censorship game in the US is strong, but not that strong. I read a book about propaganda written by someone who wrote it for the US in WW2, and he was able to explicitly state (a chapter or two in) that he only gave examples of axis propaganda in the book because his side's were classified, but to remember that everyone did the same things.

(he did leave for a footnote that for examples of pornographic propaganda one would have to visit archives at an address in Washington DC; guess that'd be the 1950's talking)

If the US is conducting psy-ops on social media against adversaries, wouldn't they use media companies popular in that country? And, if that were Twitter, wouldn't the language barrier still make it relatively harder for Twitter (a company staffed predominantly by English-speaking employees) to investigate?

Additionally, does it merit discriminating between offensive and defensive psy-ops? Might there, for example, be government-controled bots injecting wholesomememes content into the feeds of depressed government employees? Does Twitter have a different obligation in that context?

> wouldn't the language barrier still make it relatively harder for Twitter (a company staffed predominantly by English-speaking employees) to investigate

That doesn't seem to have stopped them from banning accounts posting in Chinese.

> government-controled bots injecting wholesomememes content into the feeds of depressed government employees

That would still be "coordinated inauthentic behavior".

There’s fake ones like the @antifa_us account that was run by a white nationalist organization. It was promoting violence against white neighborhoods and was amplified by the Trump campaign and conservative media.

It has been verified by reputable media that twitter handle @antifa_us was a fake antifa poster organized by white nationalist group Evropa ([1],[2]). It's impressive there has been so much false action that is immediately discovered during this time.

In the small towns around the Seattle area there have been multiple false claims of antifa planned attacks, leading to lots of people pulling out their guns. Here's one where someone visiting from out of town took pics of kids walking around and made wild accusations that they were 'outsiders' who were threatening - yet he was the outsider who was making up threats. [3]

1. https://www.nbcnews.com/tech/security/twitter-takes-down-was...

2. https://www.redstate.com/elizabeth-vaughn/2020/06/01/antifa_...

3. https://www.issaquahreporter.com/news/how-a-man-from-califor...

>23,750 accounts that comprise the core of the network, e.g. the highly engaged core network. Approximately 150,000 accounts that were designed to boost this content, e.g. the amplifiers.

That's a lot of 五毛

Since "五毛” is used everywhere for netizens to verbally attack anyone who speaks positive towards CCP and Chinese government, I'd say this seems an inappropriate use of language.

Instead, we probably should label this group as the “government sponsored online moderator” "网评员”. This is more revealing of what actually is happening, I.e. Chinese government hires people to influence and monitor online communities.

On the other hand, there are many people who does think CCP and Chinese government is good overall, and support them. Which I do not think it's fair to label them as "五毛”. In case you find this unbelievable, I'll risk being further misinterpreted, by citing the die hard trump supporters being labeled "white supremacists", which I believe is harming the unity of this country.

Can someone add their commentary as to why this person is getting downvoted for making perfectly reasonable comments in this thread?

Here's another example: https://news.ycombinator.com/item?id=23495221

Read my own comments to understand that I'm as anti-CCP as anyone on HN and extremely cynical about China, but let's not just bury comments for no reason.

ardit33 8 months ago [flagged]

^ it seems that HN needs to do some garbage collection, and remove the state linked accounts here as well

It breaks the site guidelines to insinuate this with no objective basis. Please read https://news.ycombinator.com/newsguidelines.html and stick to the rules when posting here.

Another user expressing a view you dislike is evidence of nothing but that a topic is divisive. Accusing, attacking, and insinuating otherwise about people is one of the most popular internet sports, but it's poison, so we don't allow it here.

I don't have any windows into state activity but we have extensive experience with internet behavior on HN, and I can tell you for certain that the vast majority of these accusations are 100% pure projection. People run across a few data points and connect the dots with their imagination, usually to concoct a sinister picture to explain what they notice and dislike. An internet forum is a massively multiplayer Rorschach test. That's why accusations of abuse need some objective basis or some form of evidence or...something, rather than nothing at all. It's all too easy to fire off rounds of these cheap shots without considering who you may be wounding (because that's what unjust accusation does) or what it does to the community.

It seems that if the distance between the other person's view and our own is too great, it's somehow too much of a stretch to believe that they could possibly be sincere, so we turn them into enemies, manipulators, astroturfers or spies. Now just imagine what this does to someone foreign (say, a 22-year-old Ph.D. student studying in another country and language) when they show up and they get hounded in this way, accused of being a foreign agent or a communist shill or whatever it is—simply because they come from a different background and speak from that perspective.

Is that the kind of community we want to be? how we want to treat others? Of course not; it's ignorant and indecent. What would you say about a small country town that treated outsiders like that? Every HN user would have a lot of choice language to label that sort of behavior in others, and would feel utterly superior to it. Yet we're constructed the same way and do the same thing, and not only do we not see this in ourselves, we're certain that we're right and noble for doing it.

I don't mean to pick on you personally; like I said, it's popular. But it's harmful and it comes from someplace dark in human nature, so please don't go there here. This isn't theoretical—people have already been hounded away from HN simply because they were of a different background and tried to express their views. That's sickening. None of us wants to be a person who would do that or belong to a community that would do that, so let's take conscious care to actually not do it.

HN is a much more international community than it seems, first because HN members are living in many different countries, but also because many HN members immigrated to different countries, or their families did, not so long ago. Some are from China or have Chinese family background, but there are a lot of different backgrounds here. We can either turn that into a strength by building the capacity to hear opposing perspectives, or we can behave like a mob. Unfortunately, mob behavior is the default, because it happens unconsciously and unintentionally, and actually we feel quite innocent when doing so—so if we want any other outcome, we need to work at it. Please let's do that work.

This has come up a lot, especially in the last year or two. Here are some past comments if anyone wants more:










you're leveling a claim i might agree with, but i don't think you're doing so in an appropriate way (as in, if everybody behaved this way when similar feelings arose, we'd have a real shit community)

if you have a problem with the statement, can you address where it's inaccurate? or can you be more explicit with your own assertion in order that the rest of us can address where your inaccuracies are.

GP here, I am moderately surprised that the parent post implied I am linked with Chinese government. Usually people would be more subtle.

I was born and raised in mainland China, left to US for PhD at 22.

I gradually understand why westners show various degree of hatred towards the Chinese government, CCP, and not so rarely the Chinese people. And also see the equally ridiculous negative biased of Chinese people towards westners.

From a self-interest perspective, I try to state the less emotional charged description on news relevant to China. Because I am seeing the risk of becoming one of the scapegoat of the Sino-US conflict. For example, while I was called linking to Chinese government, I am equally possible being called "banana man" (yellow skin with white heart, you get the idea) by passionate Chinese netizens.

I am actively risking my reputation on this forum by behaving like this. As for a long time I am well aware of the genuine political and cultural biases.

But I believe this is best for my situation. I cannot just watching the hatred grow between the common people without an action. That is also hurting the my own and my family's future.

This caught-in-the-middle position is common enough on HN that I wrote about it when a similar issue came up last year: https://news.ycombinator.com/item?id=21200971. That's a long post; the part that your comment reminded me of is this:

What you're most likely getting is (for example, let's say) a Chinese-Canadian Amazon or Microsoft employee, who's been reading HN for years and is suddenly hurt and dismayed by all the aggressive anti-Chinese comments that have been showing up on the site—or (let's say) a Chinese grad student who stayed in the US, got a good job and played by the rules, and back home in China is the one holding the other side of the argument, defending the US and his American friends to his family who have been hearing nasty things about them over there.

It is a difficult position to be in, because you get attacked from both sides, and not just about something small, but in deep and painful places.

Thanks for the deep understanding shown in the posts. One correction: I am 36 years old with 2 kids born and raised in US now, was stating it was 22 years when I came to US. :)

I really appreciate that you understand the situation people like us are in. That is especially encouraging given the rapidly deteriorating situation in recent years. The great irony, however, is that, coincidentally, both China and US saw outliner leaders come into power. Like an old Chinese saying “the good fortune never come with a companion, but the ill-fortune seldom come alone” (福无双至祸不单行).

Aside from activities here, I try to proactively engage with my network in mainland China to debacle common biases towards US. My personal goal is to improve understanding.

I have changed from a heavily indoctrinated young Chinese student to become one who is able to appreciate the deep cultural convictions and altruistic intentions of American people. I have seen the same vigorous pursue of better life and unwavering self reliance from both American and Chinese people. I firmly believe the value structures of American and Chinese people are inherently compatible. The confrontation should be avoidable.

I would be hugely disappointed if US and China end up into Cold War like confrontation. That will be an unprecedented waste of human value, and squander of historical opportunity of advancing the international community to a more productive stage.

>Like an old Chinese saying “the good fortune never come with a companion, but the ill-fortune seldom come alone” (福无双至祸不单行) .

This reminded me of the Anna Karenina principle [1] (but sort of the opposite): "All happy families are alike; each unhappy family is unhappy in its own way."

[1] https://en.wikipedia.org/wiki/Anna_Karenina_principle

Well if there is an effective technical approach to do that, I'd very well support that effort.

Unfortunately, so far, only Chinese government figured out how to do that. And I think any independent individuals would not want to adopt that...

Why do you think the Chinese government needs sponsored online moderators?

I don't see a need for moderators if many people believe they are good and support them.

It certainly seems like they use them to censor people who object to them and keep themselves in power.

To suggest censoring of the phrase that calls them out is comical.

Maybe because there might exist propaganda against the Chinese government? If Twitter has evidence for Chinese propaganda on a US-based site that's inaccessible to the Chinese population, wouldn't it be reasonable to suspect that US propaganda that targets the Chinese government exists too?

The above is just my opinion, but I think we're detracting from the main issue in the parent comment, that the term "WuMao" can be considered a derogatory term. Tell me if I'm wrong, but many terms started off neutral, but turned derogatory over time. The example that comes to mind is the N-word that rhymes with aggro, I'm not even sure if it's okay to use that so I'm self-censoring.

> Maybe because there might exist propaganda against the Chinese government? If Twitter has evidence for Chinese propaganda on a US-based site that's inaccessible to the Chinese population, wouldn't it be reasonable to suspect that US propaganda that targets the Chinese government exists too?

This is making the false assumption that "propaganda" is symmetrical and equivalent. From the CCP perspective, "propaganda against the Chinese government" would be to dispute its false and self-serving account of history or to suggest the Chinese people should enjoy Western-style human rights -- basically anything that questions the ultimate authority and dominance of the party.

See: https://www.nytimes.com/2013/08/20/world/asia/chinas-new-lea... and https://www.chinafile.com/document-9-chinafile-translation.

1) Twitter is accessible by Chinese nationals in Hong Kong, Taiwan and abroad which is whom they are trying to sway.

2) US doesn't need to spread propaganda through Twitter. Trump administration will routinely just come straight out and accuse China of doing X or Y.

We're all replying to the article about the Chinese propaganda Twitter has evidence of, do you have evidence of US propaganda to submit into the discussion?

Regardless, I don't see how more wrong it is to call someone a "WuMao" than it is to call them a "trump supporter", they should both be allowed.

No I don't have evidence, and I was going to include that I don't have evidence in my comment, but didn't. I was merely stating a possibility in answer to your question. Maybe you should reply to my second paragraph instead.

Edit: I don't know about the intricacies of political correctness, but somehow "trump supporter" feels ok, maybe because it has two words and "supporter" balances "trump". Personally I think "50 cent army" isn't as bad as "WuMao" even though they mean the same thing, go figure.

The essence of my analogy was both are terms that denote the ideology of a person, something they CHOOSE, not are born into like race.

If you believed your government was good, why would you care about a label that targeted you as a supporter for them?

Calling someone a "WuMao" is saying that the person is PAID to support the Chinese government, regardless of their actual opinions, and the target of support isn't in the term itself. Calling someone a "trump supporter" doesn't include the part about getting paid.

There seems to be a negative connotation with getting paid to support something, that you're just in it for the money and don't actually mean it.

Of course there are many nuances to the whole issue of labels which I don't have the knowledge or time to go into.

I agree with the paid part adding an extra amount of bad connotation, that was an oversight in my analogy.

Regardless, being called any form of a shill should not be censored and the root of it's usage comes from a legitimate cause, especially when you're not using the phrase at a specific person, just an event like the GP did.

Using the phrase the way he did, he didn't target anyone but the CCP.

> being called any form of a shill should not be censored

It doesn't need to be censored, but calling people shills just for disagreeing makes a civilized discussion less likely.

The GP didn't call anyone a shill, he called the group of bots shills. I don't see why we're talking about the word he chose to call the bots, not the bots themselves. He didn't attack anyone, he was referencing the high number of CCP bots found by Twitter.

I thought it's well known that the government hired moderators are for propaganda purpose, right?

Random question: what's the word for propaganda in chinese? I am also chinese but can only speak with poor vocab. Google translating it yields 宣传, which is a word that I do know but it means something more like "publicity" which loses all the negative connotation of the word propaganda.

I just want to make sure it's not a common mistranslation and that people aren't talking past each other. One side is saying the government is actively lying in order to manipulate people. And the other side might be interpreting as the government working on their public image... which is uh a lot more reasonable when put that way.

Funny how language shape the way we think.

It also works the other way around: Chinese organizations with 宣传 in their name will frequently translate it as "propaganda", apparently unaware of the negative connotation it has taken on in English due to the association with "enemy propaganda".

When the Chinese government accuses someone of "actively lying in order to manipulate people", they tend to call it 谣言, i.e. "rumors".

I tried translating with Baidu and it comes up with "宣传" and "鼓吹". The latter seems more appropriate for propaganda that puts something in a positive light, but doesn't apply to other types of propaganda. So I think a good translation would have to depend on the context.

Exactly, my question essentially is, why do you think the CCP needs propaganda?

Because if many people thought they were good and supported them they could win on merit alone, not propaganda.

I don't think you need to convince me that Chinese government have inherent flaws.

And you might be wrongly thinking I am one of those supporting Chinese government and CCP here.

You can see my other reply above to see more explanation of my position.

After viewing your other reply, I do not believe you are supporting the CCP, but I also disagree with your stance.

I believe we should stand up against dictatorships and challenge them even when we have ties to them.

Being able to say anything and challenge anything is the foundation of freedom, ANY form of censorship threatens that.

People in the U.S. appreciate being able to call out their government leaders I'm sure.

Imagine if you were hauled off at night because of what you posted online.

We should all be standing with Hongkongers on this one.

tldr: the answer to government censorship is not more censorship of phrases ("五毛”) that call them out

What government doesn’t use propaganda?

Well, the article is about the discovered government propaganda by mainly the CCP.

If you have other reports about services finding bots that shilled government propaganda from other state actors you're welcome to share them.

I certainly don't like the mainstream media in the US and I do believe they peddle domestic propaganda, that doesn't mean we divulge in whataboutism and distract from this specific article we are all discussing though, definitely not without specific evidence or events.

That’s like asking why does Taylor Swift pay people to moderate her social media if many people like her.

I didn't realize Taylor Swift's social media uses bots to manipulate other people's feeds.

I figured it was a normal situation where they blocked people and tweeted for her, do you have evidence otherwise?

1,152 Kremlin bots banned sounds really underwhelming. Though on the other hand, reliably detecting their throwaway accounts must be an impossible task.

What still confuses me is how do people create so many bot/fake twitter accounts when you are required to have a phone number? Throwaway phones?

A very interesting question. Maybe authoritarian governments simply ask the telephone company — traditionally being government monopolies — to issue 50,000 new telephone numbers to allow signups? Or they hire people to stand on a busy street to offer a free can of Coke to anyone willing to receive a Twitter code on their cellphone? Or perhaps Twitter doesn’t require a telephone number in those countries?

You can literally bulk buy sims in russia if you visit certain stores and ask nicely, because there’s not much oversight and people selling sims don’t make much money, so this kind of side hustle is common.

It's really easy to get blocks of phone numbers these days, especially with the proliferation of VoIP providers. You don't need to hold on to these numbers for more than a day.

There are countless SMS services and it is a race of blacklisting the numbers they use.

It requires a google account to download the archive. Can someone summarize or cherry pick some examples? Just wondering what counts as manipulation and whether I would be able to tell.

I checked Russian archive. There's no manipulation , as far as I could check . I haven't dug through larger csv, but smaller one is all but harmless. News retweets and mild pro-Russia bias

Aw, you have to have a Google account to download the data.

Twitter migrated data to Google last year : https://news.ycombinator.com/item?id=18976150

That could explain it

Is this a joke ? I've downloaded Russian archive. Tweets there are not in any way promoting any party , they are mostly retweets of regional and statewide news agencies' posts. Some do have pro-Russia bias, but nothing raging and nothing I could label as propaganda.

According to Twitter, they were "suspended for violations of our platform manipulation policy, specifically cross-posting and amplifying content in an inauthentic, coordinated manner for political ends."

If you use a bot to retweet "regional and statewide news agencies' posts", that would fall under the description above.

It helps to report accounts to Twitter, Facebook, Instagram if you suspect them of being a politically-motivated bot.

I hope they go after the Brazilian bots. The amount of disinformation is staggering.

This is crap . IDK what's in the chinese and turkish archives, but banning russian accounts for their activity in 2010 ( grep SamantaDarko , for example ) , and containing no political tweets at all ?

Were there any "я сама крымчанка дочь офицера" accounts on their list?

No, that's the point. Imagine accounts retweeting wide spectrum of moderate GOP news sites, with an occassional joke or wit.

Being able to poke fun at oneself is one the components of my personal turing test. (but that would imply /r/totallynotrobots/ are indeed human. ERROR! ERROR! +++ATH +"(Oäç/"çä%()/ç%(/ç tape_drive_spinning.gif

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact