Bots Against US Read online

  This contrived world of fake news, non-human social media accounts and real, live humans amplifying each other was full-blown information warfare.

  It was conducted with some precision and some slap-on-the-wall tactics combined.

  There was a lot of discussion about how bots were being used in the early days of 2016. But these conversations were mostly of the “Hillary vs Bernie” or “Bernie vs Jill” variety.

  Many in the news media joked around about the bots, and this was seen as a minor, yet pretty constant state of affairs for a presidential election in the digital age.

  Bots were used heavily by both parties in the primaries, but really only got passing attention as to the importance of them.

  There was zero questioning of who was running command and control for the bots. Where were they originating from?

  Most of the discussion was about “whose bots were more effective, Hillary Clinton’s or Bernie Sanders’?”

  To the media it was like a joke, sort of “ha-ha did you see the latest bot attack?” It was actually very similar to how the media now covers President Trump’s tweeting—as if it is a joke, something to be discussed in a semi-funny way; even if the content is deadly serious and/or dangerous.

  The beginning of 2016 was a swirl of election activity. Primaries for both parties, advertising plus lots of bloviating talking heads turning red and yelling and screaming was the daily diet for most citizens in the US

  Because of the sheer number of “Bernie Bros” (Bernie Bros was the name of the super-fired-up Senator Bernie Sanders followers) bots following me, I started taking an interest in how these bots were being used. I spent some time examining them carefully.

  Then, I began to notice that for all the overt patriotism, pro-USA statements and all the American flags on their profiles, there were also signs of something else. In fact, it was something else that was sinister in nature and very worrisome.

  It was that in going through the various Bernie Bros bots following me, I started noticing that some of them, but not all, had Russian language (Cyrillic) characters either in pictures, in the content they were distributing, and even sometimes in the bio section of the account in question.

  That was weird. Why would a bunch of bots purporting to be patriotic Americans be showing Cyrillic letters?

  Even weirder, some of them showed pictures of graffiti-covered walls on their cover photos, and sometimes the writing on the walls (literally) was Cyrillic.

  Weird as this was, the other thing was constant use of poor grammar, non-English phrasing, and even how numbers were shown, with commas instead of dots (i.e., 9,5 instead of 9.5).

  While many people with limited education were supporting the candidates, these errors were about something else. Something not from the US, something we had never seen used against us. It was as if these bots were literally coming out of Russia.

  Because. They. Were. Russian.

  In fact, most of them were controlled by two different arms of Russian intelligence, the GRU and the FSB.

  There may have been even more Russian government involvement through think tanks and private groups controlled by the Kremlin as well, and the general US public probably is not aware of that, or the evidence the special counsel Robert Mueller may have pointing in this direction.

  This story done in February 2019 sums it up pretty well. http://www.citywatchla.com/index.php/important-reads-for-rss/14923-conservatives-and-berniebots-duped-by-the-russians-congrats-you-are-now-toast.

  But, again, in 2016, no one besides a few cyber- security geeks and hopefully some people in the US government were actually paying attention to what was going on.

  The Bernie Bros bots were vicious in their attacks against Hillary Clinton, and to a lesser degree, some of the Republican candidates running as well.

  They depended heavily on barraging affected accounts with accusatory tweets, posts, or tried to split the left into multiple parts. Often, they would send insults, or include a YouTube video, Instagram video or other embedded content.

  This New York Times article from 2017 does a pretty good job of explaining how many, if not most, of the Bernie Bros were actually Russians pretending to be Americans. https://www.nytimes.com/2017/09/07/us/politics/russia-facebook-twitter-election.html

  They were also highly misogynistic and cavalier about their attitudes, not just towards Hillary Clinton, but to women in general.

  This actually caused a number of members of the press and supporters of Hillary Clinton to call on Bernie Sanders to tone it down.

  But even with those complaints and concerns, the bot army moved onwards, really battling it out during the primaries for each party and sending massive amounts of tweets and Facebook posts linked to weird and/or fake YouTube videos, or blog postings, Reddit or Instagram posts.

  One social media platform was used to manipulate another, and so on and so on daily.

  Also, the bots amplified both positive and negative discussions and also disseminated content on their own as well that fed into the traditional media need for scoops and breaking stories.

  Chapter 4.

  John Podesta Phishing Email

  By this point, in late March 2016, something truly bad happened. John Podesta accidentally clicked on a phishing email, one of hundreds sent to the DNC, DNC staffers, and their friends.

  As a result of being a former Bill Clinton staffer and having cyber-security experience, suddenly I got reached out to one day by a senior staff person at the DNC. He wanted my opinion about an email that John Podesta had received. I was sworn to secrecy.

  This was purported to be a Google reset request. When the DNC person sent the email to me, three things stood out. While it had the proper language and notices, it looked weird, like a little too much information.

  When examined carefully, the Google reset link was definitely not that. In fact, it was not an actual Google link at all.

  They had used a clever link shortener to shorten long links to obfuscate the real link, a weaponized piece of malware.

  The third truly weird thing was they sent it to me after the fact. In other words, he had already clicked on the link, and the damage was done.

  Of course, in mid 2019, it is widely known that this specific email was the beginning of Wikileaks and the Trump campaign weaponizing these emails and republishing them.

  All of Podesta’s Gmail messages with hundreds of people (thousands of emails) were hacked, stolen, distributed, altered, weaponized, and re-targeted right back at us, by Russia, with the help of Guccifer 2.0, Wikileaks, DC Leaks, and Sean Hannity, Roger Stone, and several other key players, including General Michael Flynn, Donald Trump and his entire family.

  It was given to me to serve as yet another outside voice giving support to the notion that it was indeed a malicious phishing email.

  But at this point, Podesta’s emails had not been released, nor weaponized, into the type of push content that was distributed widely later in the election by Wikileaks.

  At this point in March 2016 it was just a phishing email, and no one really knew what was going to happen. But the level of tension with those in the know was so thick as to be able to be cut with a knife.

  This continued to grow as the Democrats began to prepare in full for their convention in Philadelphia, Pennsylvania.

  Also, at this point, it was not public knowledge that the DNC itself had been hacked, and for months. The DNC itself did not really know it had been hacked until roughly early May 2016.

  Wikileaks and other leaking websites controlled by the Russians like DCLeaks began the weaponization of information.

  It initially flowed from the DNC hacks around late June 2016, along with the releases of Guccifer 2.0 and then was retweeted incessantly by Donald Trump, Sean Hannity, General Flynn, General Flynn’s son, Russian bots, and the Trump campaign and their supporters.

  This was basically a daily cycle from July 2016 onwards.

  However, although saying that now, it was obvious
then that particular phishing email was not a good thing, It, in fact was a really bad thing and portended lots worse to come.

  By early June 2016, it was obvious to anyone paying attention that the Russian weaponized media program was on in full. 24 hours a day. 7 Days a week.

  This was infuriating. This was right before all the DNC hacked emails and hacked plans got “leaked” and published as weaponized media content. How was it possible that an Australian paper was calling the truth out, yet not a single US media outlet picked up on this?

  The few tiny outlets that carried similar stories were ignored. The US mainstream media was not going to let anything ruin the wall-to-wall, non-stop coverage of Donald Trump’s incessant race-baiting rallies.

  Why? Billions of dollars in ad receipts and television ratings like never before.

  In retrospect, even someone like me who did know, and was following all of this extra carefully, was not taken seriously both by members of the press and people inside the US Government.

  Could I have pushed harder?

  Yelled louder?

  Written more emails?

  Tweeted more?

  Maybe so. But I don’t think it was going to make a difference even if I had. It would have been very easy to brush me off as a partisan, former Clinton staffer, and ignore the truth.

  Actually, that is what happened. With the media, most of the big media outlets were raking in the cash from all the paid ads and relishing the upsurge in viewers.

  When I would tweet about the obvious and overt Russian connections to Trump, I would get incessantly attacked by bots calling me a “libtard,” “Clinton lover” and other fun stuff.

  This is partly why Donald Trump was able to have basically non-stop free media, i.e., wall-to-wall coverage in 2016.

  Some outside estimates have projected that he got well over $2 billion in free media during the course of the election. That is a lot of money. It was being supported by ads.

  The weaponized media did not start with the hacked emails.

  It started with the bots. The bots enabled the weaponization of the hacked emails, and the propagation of both fake information and fake news. They also enabled re-targeting information right back to us based on our interests, our likes, our dislikes, our social networks.

  The very technology that people love and use to be social became the most anti-social thing there could be. It just was not being seen that way by most people who were caught up in it.

  Normal conversations can be easy or difficult but usually have a certain cadence to them. There is a kind of normal way of speaking to one another, even those we may disagree with.

  However, once the bot armies started with hardcore weaponized information warfare using psychological operations and technology to get in our heads, this changed.

  This 2017 Recode article gives some examples of the fake ads that were used on various social media platforms by Russia. https://www.recode.net/2017/10/31/16587174/fake-ads-news-propaganda-congress-facebook-twitter-google-tech-hearing

  This Newsweek article from 2019 shows a few examples of weaponized bot driven tweets. https://www.newsweek.com/if-you-shared-one-these-tweets-during-2016-election-then-you-were-duped-786075

  For the first time I can remember, suddenly people were flipping out on each other about their candidates in ways that does not usually happen in US politics.

  Obviously, the United States has a long history of political theater, even political temper tantrums. But 2016 was the year where our normal conversations and normal ways of agreeing or disagreeing with each other found themselves under attack.

  This part of the information war attack came from the bot armies. The results of being bombarded with false, misleading, mischaracterized information specifically designed to alter our interactions with each other are still in play.

  The Cambridge Analytica/Facebook information that came out in 2017/2018, clearly showed how psychometric exploits were done on us every minute of every day.

  No one has really ever tried to put together a total database of bot-driven information spreading on all social networks, video sites, and search engines.

  Each social media company has reluctantly coughed up some metrics about fake accounts, bots, and/or specific accounts tied to Russia and the Kremlin.

  But there is no comprehensive database, and probably never will be, as much of the data has been deleted, changed, or the accounts removed.

  Also, much of the public data is only available for varying periods of time, depending on the social network, so researchers can’t really go back and piece the whole thing together either.

  Chapter 5.

  Bernie Bots All Flip to Trump in One Day

  One day in May 2016, when Trump finally cleared the path by winning enough primaries to clinch the republican nomination, I happened to be online and on Twitter, when I noticed that all of a sudden, all these Bernie Bros bots were not anymore.

  They were suddenly Trump bots. Suddenly. Within a few minutes, they all switched. I know it was probably more than a few minutes for them all to flip, but it appeared that way to me at the time.

  It definitely happened within a day or so.

  Here was another weird thing going on with these bots.

  It was freaky weird actually.

  Why would all these bots that had been supporting Bernie suddenly flip to Trump, yet still with all the Cyrillic lettering, bad English, weird numbering, and strange phrasing as before?

  So being a cyber-security geek, I decided to test my theory a bit. I tweeted something pretty innocuous about Donald Trump, using his Twitter handle to see what would happen.

  Within about 3 minutes I had over 60 responses, which is super unusual on Twitter, unless you are a rock star, or someone with a huge following. So I dove into looking at who responded to me.

  And 99% of them were bots, with canned responses, and often poor English or weird phrasing.

  So, I did a second test, this time without using Trump’s Twitter handle, just his name in a sentence.

  Again, I got slammed with responses! Not quite as many, but still in the dozens. Again, they bore the same set of hallmarks: poor English, canned responses, weird phrasing.

  All of the bots that hit me that day professed to be “honest Americans” or “Patriotic loves apple pie” and even some “democrat supporting Trump.”

  There were some great pictures of the flag wrapped around people at barbeques.

  But why would so many of these so-called “patriotic American” accounts have Cyrillic lettering scattered throughout the digital remains of their accounts?

  Why would so many of these “Americans” not be able to write well in their supposed native language?

  Because they were not Americans.

  They were Russians pretending to be us.

  At this point in the election cycle, I still had not publicly revealed that I was going to vote for Hillary Clinton, so them attacking me was a function of me mentioning Trump, and that was it. Later, once I publicly stated on Twitter that I was going to vote for Hillary, and posted a picture of her and I together taken several years earlier, I picked up my own personal crew of hate bots directed at me almost any time I talked about the election on Twitter.

  It was odd, like a having a permanent fan club. Except they hated me.

  I often envision this as me walking around with a digital smoke cloud surrounding me 24 hours a day.

  Multiply that by everyone tweeting or using Facebook, YouTube, Google search, Reddit, Instagram, and other social media/news websites during the election.

  Every time you mentioned your candidate, you would get one type of response. Every time you mentioned the other candidate, you might get another kind.

  Or sometimes a mix of bots and live people would jump in.

  Depending on what social media platform this was occurring on, it also would alter the type of response you might get.

  Often the attack bots would then make other bots start t
o attack.

  Often there was a link in the response to some weird fake YouTube “news story” or a link to yet another social media post. Like a post on Facebook saying “hey check out my post on Instagram” etc.

  We all had this.

  If you were a candidate or someone working directly for one, then you had a full storm around you 24 hours a day, not just one digital cloud.

  I found the trolls/bots had business hours. The bots were usually 24 hours a day; however there were certain accounts that only seemed to be operating during business hours. Not business hours in the United States.

  Not in Asia or Western Europe.

  No, the time zone was one big fat one: Russia.

  First the bots supported candidates like Rubio, Sanders, Cruz, possibly others, and, most definitely, Bernie Sanders and Jill Stein.

  Remember that bots are controlled by a bot controller, sometimes human, sometimes artificial intelligence.

  We all had these clouds of bot-driven information around us as soon as any device was turned on and we were engaged online.

  How quickly were the norms of conversations and thoughts upended by this false-information-filled, distortive environment?

  What role do these bots and distortive networks play in actually causing real-life changes?

  Think about this. I am not a sociologist or psychologist. However, in my time doing cyber security, a decent amount of time gets spent listening to clients’ issues and working through their fears to enable proper execution of whatever cyber issue was at hand.

  One of the most common things we see is people too embarrassed to admit they were tricked into clicking on something malicious. Or opening an unsafe email or forgetting to use a security setting.

  Translate this to political life. Translate it to your civic and voting life. A part of your daily digital menu included false information via weaponized bots designed to alter conversations, which was being done millions of people at once, in highly micro-targeted ways. How would this have affected you? You may not have known exactly what was happening, but it is quite possible there was a feeling of something being off a bit, maybe.