#Election2016: Propaganda-lytics & Weaponized Shadow Tracking
A journey into the behavioral tracking technologies of the right-wing “micro-propaganda” machine
After running a second “fake news” data quest idea past Dave Carroll, I decided to use a simple script to sequentially visit each of the 117 websites in the 🇺🇸#Election2016 “micro-propaganda machine.”
With Mozilla’s 🔦“Lightbeam,” I was able to capture the tracking technologies lurking behind the scenes within the right-wing propaganda network. In my journey with Lightbeam to the dark side, I gathered all the tracking scripts, cookies, CSS/html/xml code, tracking images, display ads, and other tracking content I could carry back.
About 200 megabytes of it.
Why, you ask? Because all of these things load onto users’ computers and phones when they visit these sites, and like the results of #Election2016 and #Brexit, probably remain a part of their lives for long after.
Here’s a primer on the general topic I’m exploring:
Today we're going to look at some of the little known details and services surrounding the world of behavior…blog.kissmetrics.com
I am interested in how behavioral targeting and retargeting is implemented by networks of propaganda and political fringe actors, because it is likely to be part of the lesser-yet-still-influential strategies used to shape the outcome of thin-margin elections. The Electronic Frontier Foundation refers to the less ethical practices of #adtech and behavioral tracking firms as semi-legal data-sharing agreements.
During my “tracker” escapade of the shadow tracking network — composed of 114 fringe/anti-science/conspiracy/propaganda sites — I indirectly connected to *474 third parties.* The supernova-like image below is the resulting “ad-tracking network” footprint that was created in less than one hour. The graph shows how close hundreds of behavioral tracking companies, sketchy advertising networks, social network APIs, “addthis” sharing buttons, 💎data mining outfits, and content delivery providers are in the lives of the people who frequent sites like:
To communicate this data more candidly, I’ve decided call this network of trackers the privacy Death Star™: what you see below is the “upside down”— the tracking shadow of the right-wing micro-propaganda network from my last Medium piece ⤵️.
[One site in the 117 did not load (404), another was blocked for malicious content, and one was removed due to being an outlier/ error.]
See what I mean? ⤵️
The Death Star (DS-1 platform), also known as the Death Star I, First Death Star, and early in its planning stages as…starwars.wikia.com
In my hour-long #adtec-xploration, I gathered close to two hundred megabytes of potential third-party tracking material on my computer: thousands of cookies, gifs, jpegs, and tiny images, lots of ?scripts, AJAX that looks like it was coded in 2008, and few “known unknowns.”
Much of the tracking content that was readily accepted by my browser from these right-wing propaganda sites was not secure. So, not only is it the privacy Death Star™ we’re talking about here, it’s the Wild [mid]West of data security.
Just like in the trailer of the soon-to-be-released Star Wars story, people’s personal info and activities are traveling unsheilded through space and time, like an X-wing fighter that’s been hit by the Death Star’s superlaser.
x `` ` ` ` ` ` 💥— — - — - — ⚫️
After my journey into the shadow of this right-wing propaganda network, I can verify that it’s the “upside down” of privacy by design. Not only is much of the content transmitted insecure, the third parties are untrustworthy. I don’t think I’ve ever seen .ad before. Is this a new top-level domain in the Republic of Advertising?
The results here are very disturbing for the millions of people who frequent these propaganda sites. For one, they are easy targets for simple hacking techniques like hijacked connections when their outdated Android phones attempt to load third-party resources like the ones shown in the list above ⤴️, as well as the known unknowns shown in the list below ⤵️.
I’m just throwing it out there, but the glaring lack of secure resources loaded by these sites could mean that some of the fake news sites in this right-wing propaganda network (e.g., thepostermail.com below) may have been designed to compromise visitors’ machines. Hey, we’re talking about propaganda and privacy here, so I’m allowed to offer at least one conspiracy in this post.
I’m going to use the data I’ve collected for this project to make the argument that this “micro-propaganda machine” is not only a source of “fake news,” political noise, and strategic misinformation, it’s weaponized, behavioral-tracking shadow tech.
I have decided to share this “stranger adtec things” dataset from my shadow raid with anyone who.is interested. Nothing secret here: this is publicly available data. However, just like non-identifying databases and email metadata that law enforcement and data analytics PR psyops firms like Cambridge Analytica use, things always get more interesting when you put everything together in the form of a network.
🗄I. The Shadow Cabinet
Moving away from the initial discussion of sci-fi folklore, data privacy and national security, the tracking data I collected had some very interesting results. The first finding was significant, yet not very insightful. At the center of the right-wing propaganda machine’s behavioral tracking shadow was: Facebook. This is not surprising as Facebook’s “like” code is added to basically every website in the galaxy. Still, it’s good to know where the center of the storm is — especially in case it begins to drift your way.
Facebook's ad targeting algorithms are about to get a new firehose of valuable and controversial personal data.www.technologyreview.com
📩 II. You’ve Got…Conspiracy Mail!
A more interesting finding in this follow-up “tracking” research relates to how propaganda might be propagated outside of Facebook/Twitter. The number of aggressive 📩 email newsletter signup overlays (see images below) I encountered during my data collection was noteworthy.
This seems to support my initial hunch that even though everyone argues about the role of Facebook and algorithms in the election outcome, the impact of plain old email is likely being overlooked. I know — not as fashionable as algorithms, right? But, on the positive side, these emails have subjects like the “Re: Exposing the Illuminati Banking & Alien Media Cartel.” Below are some examples of what I’m talking about:
🍪 III. Cookies — Definitely Not a “Sometime Food” for Propagandists
In visiting these 114 right-wing propaganda websites, my computer encountered about 3,000 potential trackers, as well as an undetermined amount of less-conspicous trackers, such as Flash-based LSO “super-cookies” and font-checking browser “fingerprinting” tech.
🤖 IV. Paranoid Android
The main purpose in doing this follow-up data dive was to get a sense of the sorts of 👀 behavior tracking and identification technology that’s being put on users’ computers and phones in this right-wing micro-propaganda network. I feel this might be the real conspiracy, because if users can’t remove these embedded tracking objects with niche tools like BetterPrivacy, Privacy Badger, and Decentraleyes, they are likely to be followed across the galaxy by this weaponized shadow indefinitely.
∞ ….. x … ⧓ ….⚫️
For users visiting these propaganda sites on mobile devices — especially those who are using phones running older versions of Android, as well as people using outdated browsers like Internet Explorer, this makes avoiding this kind of pervasive behavioral shadow-tracking even tougher.
My previous post identified IP blocking as a prominent concern in the right-wing micro-propaganda machine, so it’s worth pointing out that VPNs and proxies do little to prevent people from being tracked — especially given the number of obvious and less obvious tracking technologies I’ve found here. VPNs simply <shield> users’ originating IP address and add a thin layer of protection on open wifi networks. This does not stop them being individually targeted. Users of these propaganda sites might also opt to install add-ons like “Web of Trust,”™ privacy “bounty hunters” that sell people’s browsing histories to unsavory third parties.
Here’s a CSV of the IP/geo/registry info for all of these domains.
Election-wise, the shadow network — meaning trackers loaded from the 114 sites I visited (see image above) — included firms that collect data on voters’ places of residence; their families, friend, and employer networks; past and future travel; website searches; workplaces/employment status; religious affiliation; health and fitness activities; social media posts and comments; insurance/debt/credit profiles; and long-term purchasing habits. Basically, they use the force to protect the interests of the data rich.
The data trails left from visiting these propaganda/hoax/viral/ misinformation “fake news” sites are being used to reflect on people’s past decisions to influence the future: It’s called predictive modeling.
In other words, they are building the Death Star™ superlaser that can effectively influence people’s future decisions
📵 V. Identifying The Shadow Weaponry
Related to my interest in ID’ing *non-Facebook* sources of misinformation around #Election2016, I found one “fake news” site’s story concerning.
It’s not the “Mark Zuckerberg Found Dead In Gay Brothel With George Soros” piece in the upper right; I’m talking about the main story that talks about “Fakebook and the Deep State Illusion.” Ironically, the author is referring to the company that’s the dominant entity when it comes to tracking, monitoring, and monetizing the past, present, and future activities of every person who uses these types of websites (see the logo at the center of the “shadow network” at the top of the story):
Another interesting find was a banking conspiracy site’s main story, which calls out @CNN for being a deliberately erroneous source of fake news. It’s more than a little ironic that a site called “Secrets of the Fed” has an Illuminati reference on the its cover page right above a Facebook “like” button (with 19,000 likes, by the way). What does this mean?
👍Facebook Likes: Social Media’s Misinformation Gateway Drug?
The image shown above, as well as many of propaganda network homepages like “Jihad Watch” and “Dave Hodges Common Sense Show” clearly display how fringe/extremist political sites include the Facebook’s “👍” button front and center. You could even make the argument it’s an essential part of their brand:
What really caught my attention after collecting the tracking data on the right-wing propaganda “fake news” network is that nearly every single site — whether conspiracy, pseudoscience, hoax, anti-banking, anti-science, anti-Facebook, and anti-tracking is connected directly to the Empire, and its intergalactic Death Star’s predictive modeling superlaser.
This makes little sense to me: Many of the 114 sites amount to little more than a Wix, Squarespace, or an “index.html” file clumsily ftp’d onto a nine-dollar-a-month GoDaddy page. Many are something an 11-year old could code themselves in half an hour. It is dubious that these websites should be able to garner tens, if not hundreds of thousands, of Facebook “page likes.”
In my view, like “Voldemort” in the Harry Potter series, people who are truly worried about the Illuminati, Federal Reserve, George Soros, SR-71s, Precious Metals, and the New World Order would be scared to even say the word “Facebook” out loud in public, much less sign up for Facebook, complete every life detail on their profile, and then read content put out by Wix-template dark side fringe news sites and click on hundreds of “like” buttons.
Still, this shows that this content must be entering social media through page likes which will later be seen by certain connections (we will never know how many) on Facebook. Through “homophily” and attention-biased algorithms, 👍 fake news content might go on to spawn further confirmation bias (via group agreement) or “rage trolling” (via group disagreement).
What I’ve done in this research post is take a Star Wars-esque look at pervasive ad tracking by diving into the #Election2016 micr0-propaganda machine’s behavioral tracking shadow. Using publicly available data and open source tools, I’ve touched on a number of potential major privacy/security concerns: thousands of advanced trackers, Facebook code everywhere, and even sketchy Russian ad tech (see the list and dataset I shared earlier).
There is no question these trackers are on millions of personal and work computers, mobile phones, and even the family iPads of anyone who has ever visited these propaganda sites on these devices. So, what about “fake news?”
My previous research argued that what we are calling “fake news” is more akin to a micro-propaganda machine. This research data shows that what hides behind fake news sites is a highly coordinated behavioral tracking weaponized shadow network. The underlying tracking technology that fuels the privacy Death Star’s predictive influence modeling superlaser seems to play a part in:
A. Surveilliance of the key actors who produce propaganda — as well as all the people who respond to it
Watching the people who trigger the responses allows political strategists, data ops, and technology companies to see (and probably even initiate) social conflict in what Uber executives might refer to as “God View.” This type of omnipotent gaslighting can then be used to generate long-term data insights on millions of voters.
The more “shocking” the propaganda the micro-propaganda machine produces, the larger the social response; the more people who can be rage-trolled into fighting about things that mostly amount to coordinated distractions, the better. Why? Because all of this data can be captured and semantically mined, as well as merged with past, present and future datasets to increase the accuracy of predictive behavior modeling.
Why didn’t Clinton’s campaign work? Probably because they were being rage-trolled by “micro-propaganda machines” like the one I’ve explored in my last Medium few posts.
The data I’ve collected in this post suggests that it’s not as much about the fake news, viral hoaxes, and strategic misinformation the micro-propaganda machine creates and distributes around election times: This material helps, of course, but any group of actors will generate this type of content for you if you know how make them 😡angry enough.
My exploration of the shadow network of trackers and behavioral data collection infrastructure behind the propaganda network provides evidence that it has more to do with the behavioral data the content the propaganda machine generates. This includes—as I’ve pointed out in this post—gathering rich, detailed, and deep insights on:
- The millions of naive people who can be coerced into creating, distributing, and defending this network of “fake news” and psy-propaganda and;
- The millions of people who end of fighting about on Facebook, Twitter, in person, in the news, op-eds, news comments, on TV, in media, popular music — and especially — with their own wallets.
All of this can be coordinated with enough data and technology. Together it’s what amounts to weaponized “shadow tech.”
B. Centralized measurement and long-term behavioral tracking to build predictive models.
When behavioral data, including historical data related to content engaged with on these sites, Facebook “likes,” Twitter RTs, email newsletter sign-ups, home and work history, time spent on social media, employment and income data, purchasing history, education, frequent shopping/customer loyalty cards, travel history, and religion are combined across millions of trackers, this technology can be easily be used to generate the kind of big data needed to rapidly advance the accuracy of predictive voter models.
It’s the Empire’s final stage in constructing the privacy Death Star’s™ superlaser. Maybe a better name for it is “propaganda-lytics.” ⤵️
As I’ve discussed here, the real conspiracy is that people who frequent these sites are being emotionally manipulated on a level that’s hard to fathom:
What’s more, the covert-yet-clearly-organized propaganda machine that’s “fake news” is only half the story: The other half is the corresponding tracking network that works to capture all responses, sentiment, and personal information generated from the perpetual outrage.
How to solve it? Well, the obvious answer might be: stop fighting. That’s impossible, however, so maybe a first step might be to stop arguing and sharing politically-oriented thoughts and news content on social networking sites. Stop liking things on Facebook. The shadow is large: it doesn’t matter if posts are private or public: whether it’s a Facebook message, a group Snap, or whether it is a link shared in a private subreddit, it’s all being captured. All of it.
Don’t believe me? Scroll back up to the shadow graph and look at what’s around the center of the tracking network. Or look closely at exactly how the network of micro-propaganda actors are connected across the “fake news” micro-propaganda network.
This kind of semantic data helps political strategists and data-ops PR firms “custom-tailor” and micro-target their messages to prospective voters’ based on their life experience, social identity, and cultural background. “Fake news,” aka propaganda, helps to drown out opposing campaign messaging, but perhaps it’s become the go-to data-mining tool strategists now use to find the real gems 💎: more accurate predictive behavior models.
That’s probably why the one of the sites of conflict in the micro-propaganda network graph were informational resources like Wikipeda; it’s where modern society — especially voters and journalists — turn to when they need the “facts.”
What I’m suggesting with the data I’ve shown in the last two posts (mini data projects) is that we’ve entered a new age where the social conflict that propaganda engines generate helps shadow organizations collect data to improve their behavioral forecasting models.
It’s a feedback loop between the micro-propaganda machine and its “upside down” weaponized shadow tech counterpart: The more we argue, the more personal information, context/sentiment and rich emotional data gets captured to better target the Death Star’s superlaser.
This way, they can just sit back and push the people’s “emotional buttons” 👍❤️😆😮😥😡 instead of having to spend millions of dollars develop advertising strategies to reach people who are unlikely to ever be receptive to their campaign messages or political ideologies.
According to the collaborative work of researchers for Advertising Age, Hillary Clinton is spending over $145 million…theconservativetreehouse.com
Right wing, left-wing, and X-wing, we’re probably all part of the same game:
Looking at the trackers behind the propaganda network was something I did on a whim. It was incredible to discover that this shadow tech realm is probably the real political weapon. I plan to explore this more after analysis on the left-wing micro-propaganda component and its corresponding “shadow” of weaponized ad tech.
Meanwhile, it’s pretty dark in there, so make sure to take 🔦Lightbeam, the add-on I used to collect data for this project, and some of the anti-tracking tools I listed earlier in Mozilla’s add-on store.
💚Part III: Game 7 data analysis for this “fake news” research project ⤵️: