First Evidence That Social Bots Play a Major Role in Spreading Fake News | Hacker News Reader

First Evidence That Social Bots Play a Major Role in Spreading Fake News

technologyreview.com

70 pointstrextrex39 comments

Fake news goes further than politics. Its claws reach to everything unscientific. Those claws are -in theory- rather enormous for the amount of unscientific debate on the Internet is huge.

Take the following two examples from the alt. health community: David 'Avocado' Wolfe [1] and Dr. Joseph Mercola [2]. The former is a self proclaimed health guru with no medical degree yet he makes all kind of (dangerous) health claims. The latter's similar tho he does have a degree. The latter's article is on a website called Quackwatch. If Quackwatch is credible (I didn't verify) it could be interesting to hook into them, perhaps via an API.

Funny enough when I searched for these two (I used DDG, YMMV) in combination with the term fake news I found articles where they comment about other (supposed) fake news. For example, in one article they claimed in the headline CNN distributed fake news. I don't know if that's deliberate, but at least from a SEO PoV it seems clever of them.

Another interesting question is, could -in a future- whitelisting be a better default modus operandi than blacklisting?

[1] https://au.be.yahoo.com/lifestyle/real-life/a/34572713/david...

[2] http://www.quackwatch.org/11Ind/mercola.html

>So widespread has this become that a number of independent fact-checking organizations have emerged to establish the veracity of online information. These include snopes.com, politifact.com, and factcheck.org.

As long as it remains an assumption that these are 'independent' non-partisan fact-checking organizations, I would discard any conclusions drawn here.

Also Breitbart is highly editorialized news and commentary, but it's certainly not fake. The fact that its CEO strategized the largest political upset in recent history is very, very real.

In the context of that sentence I take independent to mean "unrelated by connection to each other" rather than "unrelated by ideology." In that context I don't see a claim that they're non-partisan.

For that matter, nor can I think of any outlet in modern days which would both a) report that information were they to find it and b) not be immediately and loudly labeled partisan by a significant portion of the internet.

many of breitbart's articles are so far removed from reality that they border on fake, and the onion is definitely fake as they exclusively publish satire. So although those sites may not be explicitly 'fake news', their inclusion in this research makes perfect sense.

Obviously "fake news" has become used for content in which truth isn't really a dominant virtue...especially when it's being overridden by "strategizing political upsets"...and especially when that's tied to just making a shitload of money manipulating people.

Traditional editorials like, say, in the National Review or NYT, may well be wrong, but at least they're trying to be right. Likewise, the "independence" of fact-checking orgs doesn't lie in being "non-partisan" so much as making speaking truly an overriding value.

(I'd concede that Breitbart in particular probably isn't a paradigm example of a fake news outlet, but it's obviously on the spectrum.)

Having said that, I suspect that Breitbart, The Onion, CNN, NYT, Buzzfeed, superPACs on both sides of the aisle et. al. do use bots to advance their social media engagement- because why wouldn't they?

There's 2 elements to this study, one which is determining what is "fake news" and the other is "how do bots influence social media engagement?".

The latter is important to study but how do we judge the organizations singled out in this study without a baseline of "real news" (for lack of better word) to compare them to?

What we can do is say, "How many stories do these various news organizations promote that are obviously and comically fictitious.

That "spirit cooking" story was a great example of a totally absurd sequence of decisions designed to create a talking point which was subsequently promoted, so it's a good point.

But the article's point is that if we start making it very hard to have fake users, suddenly these games of signal boosting become massively harder.

>But the article's point is that if we start making it very hard to have fake users, suddenly these games of signal boosting become massively harder.

I agree with this point and think that the cost/benefit analysis of allowing bot users on social media sites should be re-evaluated.

The sentence before the ones you quoted is:

> At issue is the publication of news that is false or misleading.

Initially in the election, "fake news" was a term primarily used to refer to stories and sites that were complete fabrications, created to make money from ads and for political ends. Use of the term quickly expanded to include more long-standing practitioners of distortion, like Breitbart.

The whole point of this article is that "fake news" can have real consequences so the supposed successes of Steve Bannon has nothing to do with the veracity of the publications headlines and stories.

Here's just one example of a fake Breitbart headline:

>Condemned as "fake news" by some media pundits, the Breitbart article was headlined: “Revealed: 1,000-man mob attack police, set Germany’s oldest church alight on New Year’s Eve.” However, according to local journalists, there was no mob and the St Reinold Church – which is not Germany’s oldest – did not catch fire. Local police said the night was “rather average to quiet” and the number of incidents in Dortmund on New Year’s Eve had decreased to 185, down from 421 in 2015/16. The brief fire on scaffold netting near the church was reportedly caused accidently by a wayward firework.

http://www.independent.co.uk/news/world/europe/breitbart-new...

If that's the logic you use to indict an organization as "fake news" then you could apply that example to just about every major news publisher.

My point is that conflating Breitbart with The Onion for sake of studying propaganda trends is extremely sloppy. I not trying to be an apologist for Breitbart which is pretty blatantly right-wing propaganda and has been forced to retract stories in the past.

I despise Breitbart but I agree that there's a really sad irony when readers of The New York Times, The Washington Post, and The Economist point at others and never look at themselves.

The PropOrNot scandal was a disaster, but its victims remain blissfully unaware.

They justified it because right-wing propaganda apologists often claim their outrageous deceptions are satire. For the purposes of studying the dissemination of stories that are not true, lumping the Breitbart with The Onion makes sense to me.

If maliciously misreporting a few things a week is your bar for fake news, then just about every publication out there is guilty.

http://reason.com/blog/2015/08/03/time-smears-charles-koch-i...

Whataboutism. Downvoted.

The difference is that responsible organizations wind up firing people who cross the line.

>Here's just one example of a fake Breitbart headline

But it starts to get a bit confusing when you actually look at the accusations and rebuttals, especially when you look at the fleshed out indy article [0]

Accusation : "The right-wing American website, which enjoyed a meteoric rise to prominence in 2016, claimed a 1,000-strong mob chanting “Allahu Akbar” set fire to the country’s oldest church in Dortmund."

Rebuttal :"Mr Bandermann on the other hand, in a Q&A-style response published the following day, said that from between 6.45pm and 1.30am, groups of young foreign men formed a large group of 1,000 people. He said the fire at the church only set light to netting surrounding the building and lasted just 12 minutes." ... "He said that saying Allahu Akbar is as normal as saying ‘Amen’ in church,"

So which is fake news? On the facts, the Independent agrees (although through semantics "church netting on fire, not the church" "group" not "mob") with the Breitbart article.

If they both agree on the facts, but spin them different ways, how do you determine which is "fake news"?

[0] https://www.independent.co.uk/news/world/europe/edited-1038-...

edit: They did actually issue a correction for something they got wrong in the article

"Correction: This article states St. Reinold’s Church is the oldest in Germany. We are happy to clarify that accolade belongs to the Trier Cathedral."

Replying to the deadpost:

>They've literally _never_ accidentally reported in a direction that was opposite their editorial spin.

You won't find any news outlet who will err two ways.

I love how in the minds of certain people, every single and in some cases long established news org's are, always suspect.. except for Brietbart.

The biggest challenge to branded corporate 'masthead credibility' information sources/'news' aggregators is their reliance on firms like Outbrain for advertising revenue. Outbrain packages up salacious and questionable material as click bait and runs it on sites like SFGate, supposedly as additional 'news' from a partner of SFGate. The result is deterioration of trust and respect for SFGate.

The fuzzy term 'fake news' is fraught with problems. We've been lied to so many times in the western world via 'official' sources many people simply don't trust the mainstream media any more. Apparently more people were watching Yogi Bear reruns than CNN in the US a few weeks ago during prime time.

The joy of the internet is our ability to consume information from a wide variety of sources, triangulate across many ideas, reporting and opinions to make up our own minds. The old model of having a small number of 'news outlets', such as the four commercial TV channels and one main 'news' broadcast a night that worked so well during the Vietnam war era is long gone.

There seems to be a hankering for regulated, rubber stamped 'news' and credibility checks, which I think is profoundly undemocratic and against the principles of free speech.

> Apparently more people were watching Yogi Bear reruns than CNN in the US a few weeks ago during prime time.

lol fake news much? You got that from Sean Hannity's tweets. Cable news are reporting huge viewership

http://variety.com/2017/tv/news/cable-news-ratings-cnn-fox-n...

http://www.washingtonexaminer.com/sean-hannity-mocks-reruns-...

I read that here on right wing financial site ZeroHedge. http://www.zerohedge.com/news/2017-07-06/cnns-rating-collaps...

Sean Hannity's tweets - along with any other TV bobble head of either political persuasion - are not something I ever pay attention too.

I try and read a cross section of perspectives across the political spectrum and draw my own conclusions from that...

Reading a variety of garbage that doesn't share their sources or even provide a ledger for what their data means isn't going to give you good conclusions. Including any of those outlets is a mistake.

Everything is relative. In my social circles virtually no one watches broadcast or cable TV any more. The ad industry relies on Neilsen ratings to say everything is wonderful in the world of passive consumption, but do you trust their stats? http://www.pajiba.com/think_pieces/ineffective-nielsen-ratin... Your dismissive idea about not bothering to 'read a variety of garbage' implies you will only read rubber stamped, verified information. The question there, as we all seek versions of the truth, is who is doing the verifying and what is their agenda... This is why the internet is such a wonderful thing, don't be fenced in by pre chewed and spun information...

If you don't trust Nielsen ratings, why are you reading things that only show you a sliver of that data and use it to make attacks?

Reading multiple sources doesn't matter if they're all trash, you'll read multiple biased accounts and then agree with the one conforming to your own bias. I don't read "rubber stamped information," I read things that can source their data and check the sources to make sure they aren't lying to me.

Your link's only source is another article that links to themselves multiple times before providing any source, a bad sign, then the source is provided without context. It basically says "lower on this list is worse therefore CNN sucks."

'sources' is the key word here...I'm not attacking anything.

Trusting such content and repeating their claims is attacking what they want you to.

It is way too simplistic to decree certain sites fake or not, and I would have hoped than an article from MIT would acknowledge this instead of reinforcing this "fake news" meme.

If you've heard about yellow journalism as well as Project Mockingbird, then you can appreciate the fact that news is a tool to entertain and manufacture consent and occasionally inform. All of it lies on a spectrum between fake and not fake. Even The Onion which is intentionally satirical, typically has a nugget of truth in each of his stories, which is why it's good satire.

If you're not already convinced how ridiculous this term is, consider for a moment: How do you prove if a certain publication is "fake news" or not? If a "not fake news" mainstream news source publishes a single story which is later corrected or retracted, is the entire publication forever labeled "fake news" or is a certain amount of fake stories required for a publication as a whole to be considered fake?

Journalists have been making mistakes since the beginning of journalism but if you're paying attention you know there's a new class of web site that's qualitatively different from journalism. It doesn't make an attempt to do anything but spread transparent lies. And it's effective because it's aimed at a specific audience that's susceptible to believing what it wants to believe without asking questions, regardless of the transparency of the lie to everyone else. There's a specific reason the term "fake news" was just recently coined. And there's a reason why some are trying to pervert the term by applying it to anything that's mistaken. Because then it applies to everyone instead of calling attention to those who are specifically benefiting from it.

I hear where you're coming from, and it could very well just be that the term appeared to help explain why Hillary lost the election. However, I believe the term can easily be used to silence independent news sources that are telling inconvenient truths. The motive for this could be that the proliferation of new media news sources (social media, blogs, youtube) is a threat to mainstream establishment news sources' bottom line, as well as their immense power in shaping public dialogue.

Distribution of propaganda by bots is at least sort-of democratic. I wish there was more analysis of the automated suppression of news by the various platforms (twitter, facebook et al).

It's not democratic at all. It's specifically the use of resources to impersonate votes.

Ok, not democratic in that each person doesn't have an equal voice, but democratic in that anyone could buy some twitter botnet time and get their own message out.

It's not traditional to raft the notion of democracy to "how much money you have to rent a resource."

In practice, sure. That can happen. But we tend to universally awknowledge that it dilutes the intent of democracy, which is to keep a government honest to the people it governs. It seems like, given the astronomical differences in wealth between individuals in this world, this is precisely the kind of thing where one person can use resources to dilute the power of many individuals.

Would have liked to have seen a concrete example of where it worked and walked the reader through a simple demonstration. At some point in the article you realize it's all just a bunch of hand-waving.

Paper would be better titled as "Examination of Online News Amplification through Social Bots".

There are no multiple-bullet itemized criteria on what makes "Fake News" in the whitepaper.

Some of the listed websites makes uses of only links to other websites, also of dubious nature. Perhaps, that's the criteria.

>Fake news were used to manipulate the recent election

I see this being repeated often, but very little evidence behind it.

arxiv link to the underlying research: https://arxiv.org/abs/1707.07592

FoxNews and Brietbart are 2 sides of the same coin. They are right/far-right propaganda outlets, who's only aim is political discord. They will use fake or blown out of proportion or Pumped up editorial "news" segments to spread their message to serve an agenda. These bots help their message spread. Any guess who's agenda they serve?

Maybe it's justified. There's some big fear mongering and public shaming by the MSM against conservatives, especially lately.

https://imgur.com/a/c1RnG

https://www.reddit.com/r/The_Donald/comments/6rantp/does_any...

Spare me "the Donald" reddit thread. You dont think in light of certain facts thats conservatives and fascists masquerading as conservatives getting heat is justified?

Reply on news.ycombinator.com