OpenAI: Cambodian scam centers used ChatGPT to lure Indian nationals, conduct investment fraud

OpenAI said it disrupted ChatGPT use by cyber scam centers in Cambodia that integrated the chatbot into investment fraud schemes and human trafficking operations.

The scammers used ChatGPT to enable investment, romance, gambling, and law enforcement impersonation scams while also creating fake job ads aimed at people in India. OpenAI was tipped off to the scheme by officials at WhatsApp and said it banned the accounts associated with the activity. 

“In our investigations we routinely observe actors moving between scam types, or combining multiple scam techniques within a single operation,” OpenAI said in a blog post. 

“The network used our models to create and support the operation of fake online personas, generate and translate messages sent to scam targets, create promotional content for their fraudulent schemes, and assist with day-to-day operations.”

The activity was traced back to Poipet, a city in Cambodia with deep ties to scam operations run by Chinese criminal gangs. 

The scammers used ChatGPT to create fake images of passports, legal notices, stock purchase confirmations or gambling platform pages. 

The people behind the scam center also used ChatGPT to help run the operation – creating internal announcements, translating messages between members of the staff and creating recruitment fliers that could be shared in India. The fliers made fake promises of free flights, accommodations, work visas and more. 

“Users maintained records of employee debts, salary deductions, disciplinary fines, and loan repayments, and translated discussions about immigration status, work permits, visa overstays, and recruitment incentives,” OpenAI said. 

“Some conversations also referenced apparent detention, escape attempts, and potential criminal liability for people who had been trafficked and forced to work in scam operations.”

OpenAI was unable to calculate the financial losses associated with the scheme but said the operation “may have interacted with hundreds of targets across multiple scam types” and several conversations referenced thousands lost by victims. 

Scam centers continue to proliferate across Southeast Asia despite U.S. and Chinese law enforcement action that disrupted dozens of scam centers earlier this year. Experts have also raised concerns that the scam centers are expanding to South Asia and Africa. 

INTERPOL published a study on Monday that found more than half of all cybercrime across Africa now leverages artificial intelligence in some way. 

The agency surveyed 36 member states and said law enforcement believes AI is “now used to automate every stage of the attack lifecycle: from reconnaissance and phishing to extortion and evasion.”

AI-generated audio and video was used by scammers to impersonate government officials, with one incident in Uganda causing more than $2 million in losses after a public figure’s image was used to promote a fraudulent investment scheme.