How Cybercriminals are Using ChatGPT

Overview

OpenAI has exposed how cybercriminal networks are leveraging ChatGPT to run large‑scale scam operations. These groups are using AI to craft convincing messages, fake identities, and fraudulent websites, making scams more personal, organized, and harder to detect. The disrupted campaign involved romance scams, investment fraud, gambling schemes, and law‑enforcement impersonation, targeting victims across platforms like WhatsApp and Telegram.

How ChatGPT Was Misused

• Fake personas: Criminals built dating profiles, investment “experts,” gambling representatives, and false law‑enforcement identities.
• Generated content: ChatGPT produced promotional material, translated messages, and managed day‑to‑day scam operations.
• Forged documents: AI‑generated images were used to create fake passports, legal notices, stock confirmations, and gambling interfaces.
• Scam progression: Contact victims → build trust → demand money.

Types of Scams Observed

• Romance scams: Friendly conversations evolved into “investment opportunities” in cryptocurrency or spot‑gold trading.
• Investment fraud: Victims promised guaranteed returns or risk‑free investments — classic red flags.
• Gambling scams: Fake platforms claimed victims had won bonuses but demanded deposits or activation fees.
• Law‑enforcement impersonation: Fraudsters posed as officials, pressuring victims to pay fabricated fines.

Organized Crime & Forced Labor

OpenAI’s investigation uncovered signs of human trafficking and forced criminality:

• Job ads in Poipet, Cambodia, offering “chatter” roles with flights, meals, and visas.
• Internal material referencing debts, salary deductions, fines, and immigration issues.
• Conversations about detention and escape attempts, suggesting exploitation of workers inside scam compounds.

This aligns with broader reporting on Southeast Asia scam centers, where fraud, money laundering, and trafficking intersect.

Impact

• Hundreds of victims may have been contacted.
• Individual losses reportedly reached thousands of dollars.
• Financial impact remains unverified, but the scale shows how AI‑powered scams can reach global audiences quickly.

Defensive Guidance

• Be cautious when strangers raise investments, ask for secrecy, or promise guaranteed returns.
• Stop payments immediately if suspicious.
• Preserve evidence: Save chats, screenshots, and transaction records.
• Report incidents to platforms and local law enforcement.
• Verify claims independently: Don’t act on links or contacts provided by the sender.

Expert in the Cloud Insight

This case shows how AI can amplify traditional fraud models like romance scams and pig‑butchering. By automating persona creation and content generation, criminals scale deception while blending into legitimate online interactions. For defenders, the lesson is clear: AI misuse is now part of the fraud landscape, requiring stronger verification, awareness training, and cross‑industry collaboration to disrupt these networks.