How Trolls Hack Twitter to Silence the Rest of Us
Hello, my name is Patrick S. Tomlinson (formerly @stealthygeek and @patricktomlin1, both of which are now suspended). Some of you know me as a science-fiction author. More of you probably know me as that one guy who wrote the abortion thought experiment that went viral last October
In today’s social media environment, being politically outspoken and possessing any measure of reach and influence paints a target on anyone’s back. Media figures, celebrities, and creatives are all too familiar with the firehose of targeted harassment and trolling that comes with the territory of daring to voice an opinion, especially women.
Over the years, I’ve received innumerable suicide demands, several dozen death threats, had my private information doxxed five times, a DDoS attack on my website, and on one memorable occasion, a white supremacist came to my door hoping to intimidate me, only to be educated in my enthusiasm for firearms. Don’t judge a book, guys.
I’m hardly alone.
Trolls on the Loose
Trolling and online harassment is nothing new, but in the years since GamerGate, it’s taken on a more organized and frightening tone. Threats of physical and sexual violence have grown beyond cyberspace and entered the “real world.”. The practice of “Swatting,” wherein online trolls use a person’s doxxed address to make a false report of a hostage situation or other deadly threat to local police has already resulted in the death of an innocent man in Wichita, Kansas over a Call of Duty game session.
Many celebrities have been forced from various platforms by the vitriol of toxic fandoms, perhaps most famously Daisey Ridley, John Boyega, and Kelly Marie Tran of the rekindled Star Wars series bowing out of Twitter because mouth-breathing, waifu pillow-hugging, straight white man children couldn’t cope with a female Jedi, a black hero, and an Asian female character who wasn’t deliberately sexualized to appeal to their adolescent fantasies.
Other high-profile individuals have weathered these storms and come out all the stronger. People like Brianna Wu, one of the first women targeted by the GamerGate brigade pushed through the death threats and people literally live-streaming driving to her house. Today, she has eighty-thousand Twitter followers and is gearing up for her second campaign to join the House of Representatives in 2020 on a platform that rightly emphasizes her experience in tech and how it relates to security and privacy in the Digital Age (although, Brianna, if you’re reading this, gun insurance is still a dumb idea).
In response to these online terror campaigns, almost all social media platforms have either adopted or drastically expanded their reporting systems with the hopes that their vast communities would be able to police themselves and remove bad actors before another crisis occurred.
The problem is, while their intentions were laudable, the sheer, insurmountable volume of traffic on sites like Twitter, Facebook, and Instagram meant these reporting systems are largely on-your-honor arrangements on the user side of things, and almost entirely automated on the in-house side. It’s up to users to report tweets or posts that may or may not violate an arbitrary standard set by the platform’s Terms of Service, which range from the wild west to draconian. Once reported, it’s no longer eyeballs, but algorithms viewing the posts and making the call on what is and is not appropriate.
And while these algorithms are quite good at identifying offending key words or phrases, they’re severely limited when it comes to understanding context, allowing the cleverer trolls and harassers loopholes you could drive a windowless panel van through. For all its promise, AI today is about as bright as a jar of pig knuckles.
But, more worryingly still, is the opportunity for abuse of the reporting system itself. Trolls use the very weakness of the algorithms’ limited reasoning to launch coordinated mass false reporting attacks against their critics and targets, often organized in secret on another platform. These attacks work similarly to a DDoS attack on a website, swarming the software with a multitude of false reports, often generated by multiple sock-puppet accounts, in order to present the illusion of public outrage or offense over a tweet or post where none exists, triggering an automated punitive response by the platform’s enforcement algorithms.
This weaponization of the reporting system by bad faith actors contributes to harassment and increases the toxicity of online communities by forcing out the trolls’ most effective critics, making these platforms even more dangerous for genuine users, not less.
The Worst Sort of Attention
Recently, exactly this situation happened to me. On Friday the 14th, I was informed my verified twitter account with more than thirty-eight thousand followers had been suspended, permanently, for “targeted harassment,” consisting of a single tweet directed at a pro-Trump troll account spreading victim-blaming and propaganda in the wake of Trump’s conspiracy theory regarding the three-thousand Americans who died in Puerto Rico following Hurricane Maria.
Now, I freely admit that my response was crass and provocative. It was typed in genuine anger at someone trying to legitimize Trump’s pathetic attempt to shift the blame for his disastrously incompetent response to what became the worst natural disaster in our nation’s history onto his victims. Yes, his victims, as the vast majority of the people lost to Maria died weeks or months later due to completely preventable causes related to the absence of electricity, clean water, food, and medicine. All things that a competently run relief effort would have provided in time. Maria killed dozens. Trump killed thousands.
But while my response was obviously tactless, it was just as obviously hyperbolic and metaphorical. It doesn’t violate the ToS in any meaningful way. So what happened? For that answer, we have to go back to more than a week before the suspension to a random reddit thread.
Fair warning, in the screenshots and tweets to come, there will be a prodigious amount of homophobic language used, because much like the grunts of more primitive primates, anti-gay slurs seem to be the only way these troglodytes can communicate with each other. Anyway…
…as stupid as this is going to sound, the whole issue came about because people on an Opie and Anthony subreddit thread were mad I said I’ve never personally found Norm Macdonald funny. Random, I know, but here we are. From then on, my account was deliberately targeted for harassment and mass reporting attacks, including at least four fake accounts all run by the user on the above thread who bragged about being “on my sixth account.”
The “Dead Roe” theme, incidentally, relates to another Twitter user, Joe Cumia, who had his account suspended for a variety of reasons. This particular troll reveled in attacking him, and started these accounts with their corpse themes to mock Cumia’s mother who passed away earlier this year. Yes, you read that right, someone started multiple twitter accounts to taunt another human over his mother’s death. That’s the level of depravity we’re dealing with here.
Which relates to the next “tweet” that was sent via my website.
This tweet is a fake, turns out it never existed. It was photoshopped and sent to me and several others by the same trolls who got my account suspended in the hopes I would falsely attribute it to Mr. Cumia and go on the attack against him. And while a quick perusal of Mr. Cumia’s public statements and behavior prove he and I are already natural enemies, he is innocent of this. It was just another component of the troll’s targeted harassment campaign. A campaign they celebrated publicly only moments after their false attack succeeded in suspending my twitter account.
As you can see, it is hardly credible to believe these fine gentlemen were concerned with making the platform safer for other users. Instead, their purpose was to silence one of their ideological opponents so their white supremacy and homophobia could spread unimpeded. And Twitter’s reporting algorithms rewarded their efforts. But it doesn’t end there. Cyberbullying of this nature is all-encompassing. Trolls dedicate hours of research to find their targets on every platform and swarm them from every flank, hoping to throw people off balance and intimidate them into overreaction or withdrawal.
For my part, this group of cyberbullies continued to press their attack on my YouTube channel…
…on Twitter itself with an impersonation account that stole vacation pictures from my Instagram in order to falsely accuse me of pedophilia…
…and with fake 1 Star ratings on my most recent book’s Amazon reviews.
And before anyone accuses me of sour grapes over bad reviews, just stop and look at the 1 Stars. They all came in two waves of 48 hours, which just happened to coincide with me calling out white supremacists on social media. None of them are verified purchases, and none of them contain any details that would indicate they’d actually read the book. If you want to see what an average distribution of unmolested reviews looks like, just look at ratings of the same book on Goodreads and Audible.com.
Amazon is only the most visible and popular of the various rating systems, so it’s the trolls’ most obvious target for this sort of attack. Amazon knows fake reviews are a problem, but outside of high-profile examples like Secretary Clinton’s most recent book WHAT HAPPENED, they’ve done virtually nothing to defend the integrity of arguably the world’s most influential literary ratings system, again relying on pig-ignorant algorithms to do the heavy lifting of deciding which reviews are genuine.
And yes, I realize I’ve opened my Goodreads and Audible ratings up to attack from the same people who went after my Amazon ratings, because this post will certainly get shared around among them. They have nothing else to do with themselves. But when they go after my books on those platforms, it will only prove the point. These people are not brilliant tacticians. They continue to believe Reddit threads are somehow invisible to the public, for example.
In review, my Twitter account was suspended because a computer program was tricked into believing I was trying to silence someone’s voice by a pack of abusive, homophobic trolls whose transparent and admitted goal was to silence my voice. If they did it to me, they can do it to anyone. And remember, as bad as all this was, it’s not half as awful as things said and done to women on social media who dare to speak out. By now, many of you are probably wondering what you can do to protect yourselves from this sort of overwhelming, organized attack from deliberately malicious and dishonest people. I have some suggestions.
First off, don’t put all your eggs in one social media basket. In retrospect, this was my biggest mistake. Twitter was far and away my most successful platform, so that’s where I focused the majority of my time and attention growing my audience. But that left my social media presence exposed to a single point of failure. In recent months, I’ve started trying to diversify my web presence beyond Twitter into other forms and media, such as Instagram and launching a YouTube channel with original content.
But these efforts turned out to be a beat too late, and now I’m starting over basically from scratch. My best advice, especially for aspiring creatives, is to use your social media presence to drive your audience to a platform you alone control. In my case, it’s my author website, blog, and an email subscription list that should be going live alongside this article (Subscribe!). That should be the home base for everything you do. With it, you will build up a core group of fans and supporters that can’t be taken away by any troll or algorithm.
As far as what Twitter, Facebook, Instagram, etc can do, I’ve got suggestions there, too. First, appeals should be heard before punitive actions are taken, not after. Twitter’s system of offering either a 12-hour lock or an appeal process that can take days incentivizes users to just take the lock instead of fighting back, even when in many cases the reported tweet is utter nonsense. I fell for it on a couple of occasions when I knew my tweet was in the clear, but didn’t want to be cut off from the service for an unknowable number of days while my appeal worked its way through the system. Twitter’s been my primary source of news and networking for many years now. It was just easier to delete the ‘offending’ tweet and log off for half a day than to fight back. That shortsightedness came back to bite me, as each of those locks counted against my account in the most recent suspension.
Secondly, when it comes to permanent suspensions, that’s not something that should be done without direct human involvement. Automated systems just aren’t capable of weighing all of the various inputs and context in this dynamic, multi-platform social media environment and rendering a fair decision that takes not only the words, but the intentions and motivations of all the various players into account.
Only human beings can accomplish that, ideally several of them in committee to account for the natural biases and passions of their compatriots. Without such measures, you leave yourselves and your tens of millions of users beholden to the tyranny of a loud, unscrupulous minority of trolls who use your well-intentioned rules as a cudgel to beat the decent majority into silence or submission.
And Jack, I hate to say it, but do you want Nazis? Because that’s how you get Nazis.
As for the obsessive cyberstalkers who just combed through all 2300 words of that looking for some new angle to attack me, well…
Well, they’ve struck again. This time they got Chuck Wendig locked up for five hours. Also, my new Twitter account, @patrickstomlin1, has been permanently suspended as well for evading a ban, which I knew was coming eventually, but I had hoped to hold out just a little longer. Really, Twitter is an amazing platform run by either morons or people desperate to avoid any sort of accountability for what they’ve created. Not mutually exclusive, I suppose. Anyway, it’s more important than ever for you to subscribe to my email list in the box above. I promise a minimum of Viagra promotions.