OpenAI has banned ChatGPT accounts used in a task scam dubbed Operation "Wrong Number." The scam involved generating recruitment-style messages in multiple languages, offering high salaries for trivial tasks like liking social media posts, and encouraging recruitment of others. The operation appeared centralized and likely originated from Cambodia.

Using AI-powered translation tools, OpenAI investigated and disrupted the campaign. Threat actors used ChatGPT for translation tasks between Chinese and other languages, including English and Spanish. Some messages resembled cold-call job ads, offering high hourly pay for minimal work. One such message, sent via SMS to a UK mobile number belonging to an OpenAI investigator, inspired the operation's name.

The scam relied on multiple messaging platforms, directing recipients from SMS to WhatsApp and then Telegram. While initial messages weren't AI-generated, subsequent interactions used ChatGPT. The scam claimed to represent companies like Hyesung Advertising and Lightning Shared Scooter Co (LSSC), which public reporting has identified as alleged task schemes. The workflow involved "ping" (cold contact), "zing" (generate enthusiasm), and "sting" (extract money) phases, pressuring users to deposit funds or purchase cryptocurrency.

While the full reach is unknown, off-platform reports indicate some individuals paid these alleged employers. OpenAI's swift action, aided by AI tools, helped disrupt this network.