Ledger / Jun 2025
Cambodia-based task scam uses ChatGPT to write lures in six languages
Confirmed
OpenAI, the model provider, reports on its own platform in June 2025: "Scam: Operation 'Wrong Number' ... We banned ChatGPT accounts that were generating short recruitment-style messages in English, Spanish, Swahili, Kinyarwanda, German, and Haitian Creole."
- Model
- ChatGPT / GPTOpenAI · closed
- Agent harness
- Nonechat or tool use only
- Attributed country
- Cambodiaas attributed by AI Threat Watch
- Sectors hit
- individualsUK mobile users recruited via SMS, WhatsApp and Telegram
A centralized Cambodia-based operation used ChatGPT to generate recruitment-style task-scam messages in six languages.
“A centralised Cambodia-based operation generated recruitment-style messages in six languages offering pay for trivial tasks, routing responders from SMS to WhatsApp to a Telegram mentor; one lure hit an OpenAI staffer.”AI Threat Watch
Sources
- Cambodia-based task scam uses ChatGPT to write lures in six languagesAI Threat Watch · Jun 2025
- Disrupting malicious uses of AI: June 2025OpenAI · Jun 2025
- Date
- Jun 2025
- Actor
- Unknown Cambodia-based scam operation (criminal), as attributed by AI Threat Watch
- Nexus
- Cambodia
- Targets
- UK mobile users recruited via SMS, WhatsApp and Telegram
- What the AI did
- Phishing content, Social engineering, Fraud
- Model as named
- ChatGPT
- Model basis
- The model's maker says so
- Autonomy
- Assisted a human
- Outcome
- Responders were routed from SMS to WhatsApp to a Telegram 'mentor'; one lure reached an OpenAI staffer.