Inside Nasim Aghdam’s growing fixation on YouTube’s algorithm, demonetization and the grievance that culminated in the 2018 headquarters shooting.
At 12:46 p.m. on April 3, 2018, the internet’s most abstract villain suddenly acquired a street address. Nasim Aghdam, a creator convinced that YouTube had deliberately buried her work, entered the company’s San Bruno campus and opened fire on employees eating lunch. Three people were shot and another was injured while escaping. Aghdam then killed herself. Nobody she attacked had met her. To them, the algorithm was software; to her, it had become personal.
Before the Grievance, There Was Nasime Sabz
Years before the shooting, Aghdam tried on nearly every identity the early social web could accommodate. Posting as Nasime Sabz—roughly, “Green Nasim”—she made videos in several languages about vegan recipes, animal cruelty, exercise, dance and surreal comedy. One clip placed her in a sheep mask before a miserable-looking cow; another spliced stiff “ninja” moves into footage from America’s Got Talent. She described herself as the first Persian female vegan bodybuilder and briefly became a curiosity on Iranian social media.
The eccentricity was not incidental. Aghdam had constructed a world in which she was model, athlete, filmmaker and crusader, with YouTube serving as stage, employer and scoreboard. Then the scoreboard stopped behaving. Screenshots on her website purported to show roughly 300,000 views producing just ten cents in estimated revenue. She accused YouTube of age-restricting harmless fitness videos, filtering her channels and deciding who was allowed to grow.
When the Dashboard Became the Enemy
YouTube really was changing the bargain. After a year of advertiser revolts and brand-safety scandals, the company announced that channels would need 1,000 subscribers and 4,000 hours of annual watch time to earn advertising revenue. The rules took effect for existing creators in February. YouTube said most affected channels had been making almost nothing anyway—corporate reassurance that probably sounded rather different to the people being cut off.
Aghdam’s frustration therefore had a recognizable starting point. Small creators were living beneath opaque recommendation systems, inconsistent restrictions and automated decisions that could not be argued with. What separated her was the meaning she assigned to them. There is no evidence YouTube singled her out, yet ordinary platform volatility hardened into proof of persecution. The algorithm was not a co-conspirator. It was a blank wall onto which she could project one.
Eleven Hours From “Calm” to Gunfire
After Aghdam stopped answering her phone, her family reported her missing. Police found her asleep in a car in Mountain View, roughly 25 miles from YouTube’s headquarters, at around 1:40 a.m. on the day of the attack. During a 20-minute conversation, she appeared calm and cooperative. She said she had left home over family problems and was looking for work, and gave officers no indication that she intended to hurt herself or anyone else.
Her relatives later told authorities that she was furious with YouTube; accounts differ on how explicitly they warned that she might confront the company. That morning, Aghdam visited a shooting range. Hours later she entered YouTube’s campus through a parking garage and fired on strangers in an outdoor courtyard. Police found no connection between her and the victims. They were not targets as individuals, only human stand-ins for a system she believed had ruined her life.
A Machine Cannot Tell You That You Are Wrong
No public diagnosis explains Aghdam’s actions, and her family denied that she had a history of mental illness. Turning the story into “YouTube caused a breakdown” would be both convenient and false. Aghdam chose violence. The platform did not manufacture that choice, but it gave her isolation a vocabulary—views, filters, demonetization—and a faceless enemy incapable of challenging the story she was telling herself.
That is the uglier legacy of the attack. Creator platforms ask people to turn identity into inventory, then deliver life-changing judgments through dashboards and disappearing numbers. For most, the result is anger, burnout or simply quitting. In Aghdam’s hands, grievance became permission. The algorithm never pulled a trigger; three innocent people were nevertheless shot because she decided that somewhere behind the code, somebody deserved to pay.