Safeer Mohammed Koorimannil, a man trafficked to a scam center in Myanmar, impersonated a 28-year-old Singaporean woman online. In a single month, he targeted approximately 50,000 victims from at least 17 countries, according to records he smuggled out to the Associated Press.
Koorimannil chatted with over 100 people simultaneously across dozens of profiles, while supervisors used electric batons to enforce productivity. His victims included a widowed tailor in Kurdistan, soldiers in Iraq, and a dairy farmer in Georgia, among others.
The operation relied on software built with AI models from American tech companies, enabling scammers to target individuals at unprecedented speed and scale. "Everyone is a robot there," Koorimannil told AP.
This case illustrates the growing "scam economy"—a global industry where trafficked workers use advanced tools like AI and satellite internet to defraud victims across borders. The use of American AI models highlights a troubling gap between innovation and regulation.
Counter-argument: Not all AI-generated interactions are fraudulent; some tech companies argue their models have legitimate uses and that blaming tools ignores the criminal networks behind coercion. However, this case underscores the urgency of addressing how AI amplifies harm in unregulated contexts.