AI-Powered Scam Platforms 007TG and KT Exploit Trafficked Workers in Southeast Asia
Investigations by journalists and law enforcement have uncovered the use of AI-powered platforms—007TG and KT—in scam operations across Southeast Asia, including compounds where trafficked workers were forced to participate in fraudulent activities. These platforms leverage large language models and algorithmic management systems to automate scam operations, enabling criminals to scale deception while concealing their identities. The cases highlight how AI tooling intersects with human trafficking, labor exploitation, and financial fraud.
Victims of scams were targeted globally, while trafficked workers were subjected to coercion, physical abuse, and financial penalties for poor performance. The operations generated substantial proceeds, with 007TG alone linked to at least $75 million in transactions across multiple scam networks.
The Role of 007TG in Scam Networks
007TG, described as a software suite, was reportedly deployed by transnational fraud rings to orchestrate large-scale scams, including romance fraud and investment scams. Wallet records and transaction analysis connected 007TG to at least four known scam networks, though the exact revenue attributable to the platform remains unclear.
The timeline of 007TG’s documented use spans from early 2024 to December 2025, with the earliest payments linking the platform to scam networks identified in April 2024. The incident was formally recorded in the AI Incidents and Hazards Monitor (AIM) on July 15, 2026, following investigations by organizations such as C4ADS, which published detailed reports on AI-enabled scam operations. The platform’s development is allegedly tied to entities like HiSeven Pte. Ltd., though direct involvement has not been confirmed.
Forced Labor and Exploitation via the KT Platform in Myanmar
Safeer Mohammed Koorimannil, a trafficked worker, testified that he was forced to use KT to run romance scams, with the system monitoring his performance in real time. Workers faced severe consequences for failing to meet targets, including physical abuse and financial penalties, such as a $5,300 ransom demanded for Koorimannil’s release after one month of forced labor.
KT’s AI features included automated script generation, voice modulation, and performance analytics, enabling operators to track worker productivity and adjust tactics dynamically. Koorimannil’s case underscores the intersection of AI-driven fraud and human trafficking, where technology amplifies exploitation by removing the need for manual oversight of scam operations.
Mechanisms of AI-Assisted Scam Operations
Both 007TG and KT relied on large language models to generate persuasive, contextually appropriate responses during scam conversations. Additionally, algorithmic management systems tracked worker performance, assigning targets based on success rates and penalizing underperformance through physical or financial coercion.
This layer of anonymity allowed scam networks to operate with reduced risk of interruption, while the automation of repetitive tasks—such as message drafting and translation—enhanced efficiency and scalability.
Financial and Human Costs of AI-Enabled Scams
The financial impact of these operations is substantial. While 007TG was linked to at least $75 million in proceeds across multiple networks, the total revenue generated by KT and similar platforms remains unquantified. The psychological toll on victims, including emotional manipulation and financial ruin, compounds the harm.
Trafficked workers, such as Koorimannil, endured additional layers of exploitation, including physical abuse, psychological coercion, and economic deprivation. The intersection of AI, human trafficking, and financial crime highlights the need for coordinated efforts between law enforcement, technology companies, and human rights organizations to address these emerging threats.
Broader Implications for AI Security and Policy
The cases of 007TG and KT reveal how AI systems, when misused, can exacerbate existing criminal activities and create new forms of harm. The platforms’ reliance on large language models and algorithmic management demonstrates the dual-use nature of AI technology, where tools designed for legitimate purposes—such as customer service or translation—can be repurposed for fraud and exploitation. This raises critical questions about the accountability of developers, the regulation of AI-driven platforms, and the ethical responsibilities of companies in the AI industry.
Investigations into these incidents have been documented in reports such as Deceptive by Design by C4ADS, which provides a detailed analysis of AI-enabled scam operations in Southeast Asia. The findings emphasize the need for proactive measures, including monitoring of AI tooling for malicious use, collaboration between international agencies to disrupt scam networks, and support for trafficked workers to escape exploitation. The cases also serve as a cautionary example of how unchecked AI development can enable large-scale harm when deployed in unethical contexts.
Sources
- incidentdatabase.aihttps://incidentdatabase.ai/cite/1589/
- incidentdatabase.aihttps://incidentdatabase.ai/cite/1588/
Incident summary from the AI Incident Database, used under CC BY-SA 4.0.