Task scam: AI-assisted fake review jobs
OpenAI ● Covered by 2 sources
OpenAI shut down accounts, likely run out of Cambodia, that used ChatGPT to translate scam job pitches. Victims were being lured into fake 'review products for money' schemes that actually demanded upfront fees.
Based on reporting by OpenAI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Scammers keep finding new uses for AI, and this time it's translation, not generation, doing the dirty work. OpenAI says it banned a cluster of accounts tracing back to Cambodia that were using ChatGPT to convert scam messages into multiple languages, making fake job offers look legitimate to targets who spoke different tongues than the operators typing them out.
The scheme itself is an old one dressed up for the gig-economy era. Victims get approached with an offer to earn money reviewing products or apps online, easy work, flexible hours, the usual pitch. But the catch arrives once someone's hooked: they're told to pay a fee to unlock tasks or withdraw earnings, and that money just disappears. It's the classic advance-fee fraud, just running through Telegram or WhatsApp instead of email chains from a stranded prince.
What's notable here isn't the scam mechanics, which are decades old, but the role AI played as a translation layer. These operations, often tied to broader forced-labor scam compounds reported across Southeast Asia, need to reach victims across language barriers to scale internationally. A tool like ChatGPT removes that friction almost entirely, letting a Khmer-speaking operator craft convincing English, Spanish, or Vietnamese messages without hiring a translator or fumbling through machine-translated gibberish that might tip off a wary target.
OpenAI's response was to identify the behavioral pattern, bulk translation requests tied to job-scam phrasing, and pull the accounts. It's a reminder that AI safety enforcement increasingly looks less like blocking someone from asking for a bomb recipe and more like spotting industrial-scale fraud operations quietly using a chatbot as an efficiency tool. The scam didn't need AI to exist. It just got faster and more convincing because of it.
My take — AI-written commentary, not fact-checked reporting
This is the boring but real face of AI misuse: not rogue superintelligence, just scammers getting a productivity boost. I'd rather see OpenAI publish more of these takedown details regularly, because right now we're relying on their word that enforcement is even catching most of it, and that's a thin reed for something tied to actual human trafficking compounds.
Read more about this at: OpenAI
