News Summary

A significant spam campaign has been discovered, where spammers utilized OpenAI’s GPT model to send over 80,000 messages that successfully bypassed spam filters. The operation, conducted over four months, involved unique message generation for each recipient, allowing the spammers to evade detection. The spam was orchestrated by a framework known as AkiraBot, which automated the sending of messages to small- and medium-sized websites, promoting dubious SEO services. The spam messages were delivered through contact forms and live chat widgets on targeted websites. OpenAI has since revoked the spammers’ account after the breach was identified.