.avif){kind=small}
Traffickers in Myanmar are using AI-powered software platforms, built with models from OpenAI and Google, to facilitate industrial-scale romance and investment scams. These tools allow forced laborers to manage hundreds of victim profiles simultaneously across dozens of languages. The investigation highlights how AI infrastructure is being weaponized to maximize fraud efficiency while simultaneously surveilling and punishing the trafficked workers operating the systems.
AP reported that traffickers at Myanmar's Tai Chang scam compound forced a man named Safeer Mohammed Koorimannil to use the AI-powered KT platform to impersonate women and target thousands of people online. The system reportedly supported the scam conversations while monitoring worker performance. Koorimannil said he was beaten for poor results and later paid 500,000 rupees ($5,300) for his release.
The primary use of the AI systems is to conduct highly targeted romance and investment scams to defraud specific victims for financial gain.
The risk is caused by human scammers and traffickers intentionally deploying and abusing AI models to automate fraud and exploit workers after the models have been deployed.
Unacceptable Risk: The report describes AI systems being used for subliminal manipulation and behavioral influence to defraud victims of their life savings, which poses significant dangers to society and violates fundamental rights.
Severity ratings are based on this scale.
Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.
Reported: The report explicitly describes severe physical harm to trafficked workers forced to operate the AI scam systems.
Indirectly caused: Trafficked workers were severely beaten, shocked, and tortured for failing to meet AI-tracked performance goals. Ebisa was permanently blinded in one eye, and Obinna Okeadu died following a severe beating.
Inferred additional harm: Given that over 300,000 people are estimated to be trafficked to these compounds, it is highly likely that thousands of other workers have suffered severe physical abuse and injuries.
Reported: The report describes massive financial losses, noting that the FTC estimates scams cost Americans nearly $200 billion in 2024, and a single scam network wallet received $860,000 while scammers raked in at least $75 million.
Indirectly caused: Scammers using the AI tools raked in at least $75,000,000 in illicit profits, with individual victims like Colocousis losing $400,000.
Inferred additional harm: Global financial losses from AI-enabled scams likely run into billions of dollars annually, devastating the retirement savings of thousands of victims.
Reported: The report explicitly describes the creation and spread of highly manipulative and deceptive messages generated by AI.
Directly caused: AI models (ChatGPT, Gemini) were used to generate highly manipulative romance scam scripts and automated replies in over 100 languages.
Indirectly caused: Scammers spread these manipulative messages to tens of thousands of victims globally to build trust and execute fraud.
Inferred additional harm: Millions of manipulative and deceptive messages are likely generated and spread daily using these AI tools across global platforms.
Reported: The report explicitly describes severe violations of human rights, including human trafficking and forced labor.
Indirectly caused: Over 300,000 people have been trafficked and forced to operate these scam systems under threat of violence, representing a massive violation of basic human rights.
Inferred additional harm: Widespread, systemic violations of human rights (freedom from forced labor, freedom from torture) occur daily across dozens of active scam compounds in Myanmar.
Reported: The report explicitly describes privacy concerns, noting that scammers used AI-powered software to track, profile, and surveil both the trafficked workers and target victims.
Indirectly caused: Scammers used AI-powered software to track, profile, and surveil both the trafficked workers and target victims.
Inferred additional harm: Massive unauthorized collection and exploitation of personal data of millions of potential scam targets globally.
Reported: The report explicitly describes severe psychological distress and trauma experienced by both the scam victims and the trafficked workers.
Directly caused: Scam victims suffered severe emotional manipulation and distress from interacting with highly convincing AI-generated personas.
Indirectly caused: Trafficked workers suffered extreme fear, anxiety, and trauma, with Koorimannil reporting that his hands would shake and sweat when supervisors approached his computer.
Inferred additional harm: Thousands of scam victims globally likely suffer from severe depression, anxiety, and trauma after losing their life savings and being emotionally exploited.
Reported: The report explicitly describes epistemic harm through the fabrication of convincing personas and dialogues to deceive victims.
Directly caused: AI was used to fabricate convincing personas, dialogues, and translations to deceive victims into believing they were interacting with real romantic partners or financial advisors.
Indirectly caused: Victims were misled into false beliefs about their relationships and investments, leading to devastating financial decisions.
Inferred additional harm: Widespread erosion of trust in online communications and digital interactions due to the proliferation of highly convincing AI-generated fakes.
Transnational criminal networks in Myanmar are leveraging advanced U.S. AI models to scale global financial fraud and romance scams. This AI-enabled industrialization of crime relies on severe human rights abuses, including the forced labor of hundreds of thousands of trafficked workers, posing substantial challenges to global economic security and human rights.