One of the most prominent victims of the GamerGate harassment campaign took out a restraining order against their ex-partner, whose false accusations lent fire to the movement. The restraining order did nothing to meaningfully resolve the abuse, yet even if it had worked, it wouldn’t have stopped the GamerGate campaign. The campaign was built on multiple tiers of harassment across several forums that were radicalizing angry young people—mostly men—into hating their targets, obsessively stalking their online presences, and sharing rationales for abuse with one another.
While the lieutenants of GamerGate played an important role in calling targets and amplifying the less-followed members of the movement, they also needed those crowdsourced nobodies in order to make their target really feel the pain. You can’t take out a restraining order on a crowd, nor arrest them. Awful as their speech is, it is constitutional. But the ferment of that speech is what creates the basis for more overt forms of abuse, rationalizing and making it seem justified to dox and swat a target, leave a dead animal on their doorstep, stalk them and send the pictures to their parents, leave threatening messages at their door, and so on.
Thus, breaking up their network is the chief strategic goal. It is the least intrusive option that remains effective. It’s why people like Fong-Jones and Lorelei chose the targets they did. If you add speedbumps—friction—to those seeking to access a site like Kiwi Farms, you make it much harder to source the crowd. You make it harder to draw enough people in the vile hope that one among their number will be deranged enough to go the extra mile in attacking the target in more direct ways. Such networks radicalize their members, ratcheting up their emotions and furnishing them with justifications for their abuse and more besides.
Breaking up the network does not eliminate the problem, but it does ameliorate it. The harder you make it to crowdsource, the likelier it is that a particular harassment campaign will fizzle out. Kiwi Farms remains able to do harm, but it would be a mistake to suggest that its endurance on the internet means its victims have failed to hobble them. They’re weaker than they once were, there are fewer foot soldiers to recruit from, it’s harder for the fly-by-night harassers to access the site conveniently. When you winnow such extremists down to their most devoted adherents, they remain a threat, but they lack the manpower to effect harm the way they once did.
If citizenship and politics mean anything, they must include the kind of agentic organizing exercised by Kiwi Farms’ victims—to ensure that they could be more than passive victims. This is, after all, what the political theorist Hannah Arendt meant by the word “action.” That simple word, for her, meant exercising the very capacity to do something new, to change the rules, upend the board, and be unpredictable. It is, she argues, at the heart of what makes us who we are as a species—and the essence of politics worthy of the name.
Allowing Kiwi Farms to flourish would not have protected anyone anywhere in the world from the malice of authoritarians who seek to abuse power at every turn. They might have used the banning of Kiwi Farms or the Daily Stormer as a fig leaf of “precedent,” but keeping these sites online would not have stopped the censors. What would Kiwi Farms’ victims have been sacrificed for? Shall the shameless do as they please, and the decent suffer what they must?
What this experience reveals, and what is generalizable to future dilemmas of this sort, is that breaking up a harassment network remains the least intrusive option on the table. Perhaps pressuring the deep stack in this way is not optimal. The EFF is right to raise serious doubts, doubts I share. But then this key insight about the network effects of harassment campaigns means that the solution, however partial or provisional, lies in finding other ways of disrupting the networks of extremist abusers. If anyone should be left holding the short straw of pluralism, it should be them.
