Coverage of a Disinformation Operation Against 2019 EU Parliamentary Elections
labsblog.f-secure.com
labsblog.f-secure.com
As others have said, it's incredibly unlikely that a tweet is going to change your mind about anything.
Having worked both in and out of the journalism industry, I know that journalists get a lot more worked up about what people say online than everyone else does. They think it matters a lot more than it does (probably why they got into the job of saying things online in the first place). Twitter is completely irrelevant to most people. They're vaguely aware it exists, but it really doesn't matter.
In TFA, you can see each tweet getting a few hundred to a few thousand likes/retweets, and most of those identified by the analysts as bots. So each tweet is maybe reaching an audience of a couple hundred humans, at best (and it'll be the same few hundred humans for all such tweets). How many of those humans are going to change their vote because of that tweet?
It's irrelevant.
But it is dangerous. Politicians get very nervous about elections. If the politicians are persuaded that any of this matters, then they'll be more inclined to stop it mattering. And that means laws that curb free speech online, monitor communication, prevent encryption, and all the rest of the shitstorm we're facing.
In my country I've been seeing a constant campaign of subtle disinformation against opposition figures that is very powerful. You check the "user's" Twitter profile and they'll have a bunch of pro-opposition opinions on it. Then, the "user" will start to express disillusionment about certain people, asking them not to betray "our side" etc. By the end, it's a full blown attack against "traitors".
This kind of stuff is very subtle and powerful. I've seen it over and over and people will get dragged into believing this stuff and it's very damaging.
It's almost stuff straight out of the movie "Inception". You just need to plant that "seed" of doubt and it will grow in a lot people. It might just be the nature of our political climate, so it might not work everywhere though.
Don't get me wrong: I think online "propaganda" can be effective - just not on Twitter, because of its scale and primary audience.
But because Twitter is more transparent it gets analysed more. I guess you could make the argument that the same people are doing the same things on FB and Whatsapp, but it's difficult to see how those messages propagate so well.
You may not agree with it, but I really don’t think you’ll cling to the idea that they’ve been hypnotized by internet trolls.
I'm not familiar with the reasons of the outcome in France, but this whole election has been dictated by far-right, racist and highly partisan reasons in my country. When you wonder why you got a result from a huge number of people you need to analyse the data, and analysts are noticing that ...
"The League of Matteo Salvini gets a resounding 30.75% of preferences in Riace , the municipality of the Reggio area guided until a few months ago by Domenico Lucano and known throughout the world for the model of reception and integration of migrants that was practiced there."
"Even in the island of Lampedusa , at the center of the migration phenomenon, the League wins. Matteo Salvini's party gets 45.85%, more than twice the Democratic Party that launches Pietro Bartolo, the migrant doctor, among the candidates in Strasbourg."
https://translate.google.com/translate?sl=auto&tl=en&u=https...
> Doesn't it depend on how close the election was to begin with? If it's very close, you may not need to move the needle very far to change the outcome.
There are 12 points of distance between the 1st party (extreme right) and the 2nd (pro immigration). The territorial distribution of votes in favor of the extreme right is aligned with the cases of uncontrolled and non-regular immigration, cases that the population lives in the area and which have little interaction with the (old/new)media.
This doesn't mean that disinformation attempts are not happening, they're. But the outcome was predictable, and it's difficult to argue that there is a correlation with ..Twitter.
A thousand people is not enough to influence an election. A single article is not enough to influence an election. Yes, it's possible that something goes viral and gets seen a million times, but that's not what we're talking about here.
There are exceptions, of course, and advertising can be effective, but there is a lot of wasted money spent trying to persuade people of things.
We have pretty significant evidence over the last decade indicating otherwise. The RAND institute did an entire paper about it.
>If the politicians are persuaded that any of this matters, then they'll be more inclined to stop it mattering. And that means laws that curb free speech online
Pretty sure misinforming the voting public matters? Uneditorialized journalism and publishing created this issue. Don't try to pretend any law addressing this problem is somehow constraining the free speech of the public.
You are right that normal people don't read Twitter.
But Twitter users read Twitter. And there are many of them. And they amplify what each other says, conspiracy theories, telephone game, and intentional disinformation alike.
And… eventually some of that disinformation reaches critical mass and graduates from the disinformation Petri dish of twitter.com and releases spores into the mindspace of journalists, "journalists", and normal people who happen to be on Twitter sometimes.
And then, having evolved into a truth-resistant strain, that disinformation spreads through Facebook, InfoWars, well-meaning television news outlets, and television news outlets owned by Sinclair Broadcast Group, into memorable one-liners that reinforce one's own beliefs just enough to avoid application of even the hint of rational thought, which get tucked away in that special space of the mind used to store those juicy "gotchas" you're saving up for next Thanksgiving.
And come Thanksgiving, you use them – take that, Hillary-loving millennial nephew! And, you find out they're wrong, because of course they're wrong, they're ridiculous – why would Hillary Clinton be performing satanic rituals underneath a pizza place? – but it doesn't matter. You've spent the past year reinforcing your beliefs because you half-remembered that you heard somewhere a vague description that Hillary did this awful thing, and you told your friends and it reinforced their beliefs, because even if it made no sense to you, it was vague enough to make sense to them, and, well, she probably did something else terrible anyway, and in fact, your just buddy told you about how Wall Street is paying her to get rid of US borders. Sounds about right!
In terms of scope, scale, and reach you're talking about the difference between a butterfly flapping its wings, and a nuclear weapon. And the butterfly is supposed to come out on top? I think many people simply lack the empathy to understand that people can see things in very different ways without in any way being misinformed, and so they look for explanations for why the other person 'must be wrong.'
For instance two things are both true: migration can help improve the economy and overall wealth of a nation; migration can hurt workers in affected fields within a nation. Instead people who are partisan focus exclusively on one side and simply pretend the other does not exist. This leads to an inability for people to grasp with why people would see things differently. It's because they intentionally avoid seeing the negative sides of their own views, or the positive sides of views they disagree with. So it leads to people mutually arguing that the other side only believes what they do because they're being lied to and manipulated, or alternatively - that they're idiots.
[1] - https://www.independent.co.uk/news/world/americas/sam-wang-p...
> Nearly 100% of the mainstream media was against the person who ended up being elected.
Including Fox News and other right-leaning news outlets? I would suspect that for people who voted for Trump their news content was close to 100% in support of him, with negative news and polling basically either ignored, minimised or re-framed as being another sign of "the elites" being contemptuous and/or scared of Trump and by extension his supporters.
Why would you assume that people are consuming political news they don't like when every indication is that they don't? You even say this yourself later on in the post:
> Instead people who are partisan focus exclusively on one side and simply pretend the other does not exist.
When you consume content that is unabashedly partisan in nature you are far less likely to subject it to any kind of critical analysis, and even small amounts of reinforcement from other sources you align with can give enough credibility to establish a particular "fact". See pizzagate or the Q-anon conspiracy theories for instance.
In the US 2018 congressional races many of them were decided by an incredibly small numbers of voters, a tremendous amount were decided by less than 5% and quite a few were decided by margins of a fraction of a percent.
I'm literally in a car leaving for the airport so I can't verify the exact numbers, but I'm confident I'm close enough[1]: Kansas' 2nd district was decided by like .87%; Florida's US Senate Rick Scott won by .12%; Georgia 7th was .15%; Minnesota 1st was for sure .45%; Illinois' 13th district was something like .77%; New York's 27th district was like .28%. In 2016 presidential race, New Hampshire was decided by like 1500 votes. That's just a sample of the extremely close races across the country.
Now, while those races might seem concerning, it gets even more concerning when we consider smaller districts, county elections, and city elections where it isn't uncommon for seats to be decided by margins of 25 people or less.
The worrying thing about disinformation campaigns isn't that the disinformation is going to suddenly convert 90% of our population into propaganda vessels, it's our modern ability to target smaller and smaller demographics in key locations with information in ways that we previously were unable to do in cost effective ways. The concern (and it is a concern) is that the disinformation and often intentional outright lies will misinform just tiny tiny fraction of a percent of vulnerable logic challenged people in key areas where the ripple effects punch outside their weight.
And also of course how easy it is to swing local county and city elections, considering that a rather high ratio of local voters tend to be elderly people who may not have their skeptical hats on when the magic box tells them the opposing party is eating babies or whatever.
If you've spent any time having long casual conversations with many of our voters, you'd find out that many of these people have been convinced of and believe in truly outlandish things, they want to believe in fantastic explanations rather than what is usually a mundane explanation. We don't have to look far to find outlandish ideas floating around our population--just consider anti-vax and how it spread mostly organically, it was not a highly funded and highly targeted campaign.
Our people can make rational decisions and vote for their interests, provided they are given solid information and are not exposed to outright disinformation. It doesn't matter which direction people vote, as long as their decisions are not based on disinformation spread intentionally.
[1] I'll verify those numbers once I'm settled in, but I am confident they're within a reasonable accuracy for this topic. If you're still not convinced of the larger implications here, I'd be happy to list more. There were a whole shitload of key counties where elections were decided by margins of less than 10 people and quite a few were less than 5.
Yes, but multiply that by the number of targets, then by the number of messages, and then by exposures and you'll see that very small chance turn into a massive tide. That's how advertising works.
While a single tweet won't change views, the repeated exposure to a narrative will, eventually, do that. We are not talking about individual tweets, but concerted efforts to further narratives for various purposes.
I personally know people that, when provided clear evidence of entirely false and manipulative stories will justify them by saying, “well, this story may not be true, but there are a lot of stories like this that you haven’t seen which are true”.
The point is, they believe what they want to believe, and no amount of proof to the contrary will change their mind. In fact, they often just double down on their arguments.
All people are like that, just everyone for different topics. I heard so many people say "There are not enough resources on earth so that everyone can have a mobile phone" or the like, which obviously is false. Don't get me started on all the beliefs of lefts and rights in the refugee crisis in Germany. Or take "the free market solves it all" from many liberals, "capitalism is evil" from young contrarians and whatever conservative thoughts from older folks.
My point is, the problem of blind belief is not specific to the audience of those tweets, it seems human nature.
It sounds like they're looking for someone to blame for right wing candidates' success in the same way people blamed Russian bots for Trump's success because they can't believe that real people could possibly be voting for such obviously "wrong" candidates of their own free will and they must surely have been fooled into it by clever tricksters.
Since the media consumed by EU voters is so disjoint because of language barriers, this should be somewhat possible to test. If one could show a correlation between the amount of "trickery" present in media consumed by voters in a country and that country's election outcomes, you could get an answer to whether voters really are fooled into voting for right wing candidates.
Trickery works. It's why everyone keeps doing it.
Yes the "death tax" was an outright lie. But there were lots of factors which went into that margin.
Also it's important to clarify that (for the most part - except possibly on some WeChat groups) - this wasn't a foreign campaign. It was mostly old fashioned paid advertising by the Liberal party itself.
And nobody is attributing right wing success purely to bots. That's just some strawman that you invented. The issue is that is a significant factor and is causing polarisation and distrust in institutions. And you need those instutitions to work in order to solve the hardest problems e.g. mass migration, ageing population, climate change etc.
Foreign anti-western forces feed off of that, because the modern western right is essentially anti-western from the start.
No purely. But people certainly attribute it to bots or they wouldn't be concerned about bots if they believed they had no effect.
Of course this is no different to the US and the rest of the West using the media and social networks in other countries to promote their own agenda.
The difference is we're supposed to be immune from having it done to us - and of course we aren't. Because it's no different in principle from a fake advertising and promo campaign, or a fake pro-corporate stage-managed PR campaign - just larger in scope, and weaponised as opposed to financialised.
In reality troll and Tweet farms are incredibly cheap compared to other propaganda vectors, and incredibly cost effective.
At some point knowingly misleading the public for political or financial gain is going to have to become a criminal offence.
Currently consumers/voters have more protection from fake claims about a toaster than they do from fake claims made by a president or a hostile foreign power.
If you're trying to be a genuine democracy, that's a bad place to be.
You seem to believe that people can't be trusted to think for themselves and must be shielded from lies to protect them from their own stupidity. I agree on the first part, but not the second which is arrogant authoritarianism. In democracy, the voters are also the ones who experience the consequences of their decisions and them having the ultimate power over themselves is what protects them from extremes. They might hurt themselves a little bit but it's still themselves doing it, so they won't go too far and can change their mind next election if they made a mistake.
If there's something new and unique about social media that actually breaks democracy worse than anything since the ancients Greeks, then perhaps you'd have a point, but I don't see any evidence of that. I think people will adapt to false information and learn to decide what to believe, just like they currently make decisions about how much to trust any stories they hear. There's nothing wrong with having a few conspiracy theorists or people who believe a wrong version of history. They keep everyone else on their toes and we can't really be sure that it's not us who's wrong. Remember when God obviously created the world and anyone who disagreed was punished for spreading dangerous fake knowledge?
Yet people (even here on HN) refuse to look at the evidence and pass around things like the Seth Rich conspiracy as facts[1].
The same thing happened here with the Comet Pizza conspiracy theory.
Facts don't matter, and they just act as ways to polarize people.
Edit: and see how the voting goes on this and the parent comment, and on the thread I linked. The battle for facts is lost, and partisanship is all that is left.
What conclusions can we draw from this fact?
> According to Facebook, the IRA purchased over 3,500 advertisements, and the expenditures totaled (sic) approximately $100,000
I think everyone would like to know how much actual influence it had in the election but I'm not sure we will ever know definitively.
https://disinformationreport.blob.core.windows.net/disinform...
The two biggest caveats about it in my mind are that 1) all the data was provided voluntarily by the tech companies, with no indication of how they determined what was a Russian account, and 2) they don't really have any way to measure what actual impact it had on the election, or how effective the disinformation campaigns were