OpenAI Disrupts Poipet Scam Network Using ChatGPT Across Multiple Fraud Schemes
20:33 · August 5, 2026 · Hacker News AI Section
OpenAI said it disrupted a Cambodia-based scam operation that used its generative artificial intelligence (AI) chatbot ChatGPT to facilitate a wide range of investment, romance, gambling, and law enforcement impersonation schemes. To that end, it banned a coordinated network of ChatGPT accounts likely originating from Southeast Asia and operating from the city of Poipet, a region with extensive
Summary
OpenAI has taken action against a coordinated cluster of ChatGPT accounts operated from Poipet in Cambodia, a location previously linked to organized scam compounds. The accounts supported multiple overlapping fraud campaigns that combined romance, investment, gambling, and law enforcement impersonation tactics. Rather than specializing in one scheme, the operators switched between narratives depending on which approach appeared most likely to succeed with a given target.
The models were used to generate synthetic personas, draft and translate messages for victims on platforms such as WhatsApp and Telegram, produce promotional material, and handle internal administrative tasks. Examples of the latter include drafting job advertisements aimed at recruits in Bangladesh and India, recording salary deductions and debt ledgers, and preparing documents related to visas and work permits. Some generated content also aligned with recruitment patterns associated with forced labor in the region.
OpenAI conducted the investigation in cooperation with WhatsApp. The operators followed a consistent three-stage sequence of initial contact, trust-building conversation, and pressure for deposits or fictitious fines, after which victims received fabricated screenshots as proof of transactions. While the exact financial impact remains unknown, internal communications suggested hundreds of potential targets and individual losses in the thousands of dollars.
The case illustrates how generative models can be integrated into existing criminal workflows to increase throughput without requiring the operators to develop new technical skills. It also shows that frontier-model providers are now routinely monitoring for large-scale misuse that blends social-engineering content with operational support functions.
Why it matters
This article provides concrete evidence of how threat actors weaponize generative AI to scale social engineering and fraud operations. Dutch security professionals must understand these AI-augmented TTPs to enhance threat intelligence and develop robust defenses against sophisticated, AI-generated attacks.
