Threat Intelligence

OpenAI Shuts Down Cambodia-Based Scam Network Abusing ChatGPT

The Hacker News · 6 Aug 2026

Key Takeaway Train staff to verify unexpected investment offers, romantic contacts, or law enforcement requests through independent channels, since AI-generated scam messages are becoming harder to spot by tone alone.

OpenAI has disrupted a coordinated scam operation based in Poipet, Cambodia, that was using ChatGPT to support a range of fraudulent schemes. The network reportedly leveraged the AI chatbot to help craft convincing messages for investment scams, romance fraud, illegal gambling promotions, and impersonation of law enforcement officials.

The accounts, believed to originate from Southeast Asia, were identified and banned as part of OpenAI's ongoing efforts to prevent misuse of its tools by organised fraud groups. Poipet is a known hub for scam operations in the region, often linked to large-scale online fraud targeting victims worldwide, including in Australia.

This case highlights a growing trend: scammers are increasingly using generative AI tools to make their schemes more convincing and harder to detect. For Australian small businesses, this means staff and customers may encounter more polished, professional-sounding scam messages that are harder to distinguish from legitimate communications.

scam ChatGPT AI fraud

Summarised by CISO AI from The Hacker News. We link back to every original so you can read it yourself.