OpenAI banned a cluster of accounts it says were being used to power a scam operation aimed at people in the UK.
The company says the accounts likely originated in Cambodia and were using its AI tools to support scam workflows targeting UK residents, part of a takedown effort OpenAI is calling Operation Wrong Number. OpenAI has not published details on how many accounts were involved, how much money was at stake, or exactly what the AI was generating. The disclosure is one entry in a running series of abuse reports OpenAI has published over the past year, in which it names and bans clusters of accounts tied to fraud, influence campaigns, and malware development. This one stands out mainly for the geography.
Cambodia has become a well-documented base for large-scale online scam operations, some reportedly staffed by trafficked workers, that have industrialized romance and investment fraud at a scale that has drawn scrutiny from UK and international authorities. If AI tools are now doing some of the scripting and messaging work inside those operations, it likely lowers the skill and time needed to run scam workflows at volume, even as it leaves a trail in the AI provider's own logs.
Banning the accounts stops this specific cluster, not the compounds behind it, and OpenAI's report gives no indication of how quickly the operators simply moved to another account or another tool.