A variant of ChatGPT, originally crafted to aid cybercriminals, has backfired by duping potential buyers seeking access to its services.

Earlier in July, PC Mag wrote about  WormGPT, a chatbot developed from publicly available source code, which claimed to assist hackers in generating phishing messages and malware for a monthly fee. This raised concerns about the potential for generative AI to facilitate computer hacking and exacerbate cybercrime.

Ironically, it seems that the WormGPT brand has turned into a greater hazard for hackers than for the general public. Antivirus provider Kaspersky identified multiple websites purporting to provide entry to WormGPT, but instead appeared to be devised to deceive prospective customers into parting with their money, without actually providing access to WormGPT.

These websites, accessible on the public internet and through a Google search, have been adorned with official-looking details about WormGPT. Nevertheless, Kaspersky suspects that these pages are, in reality, phishing sites engineered to deceive users into divulging their credit card particulars or surrendering their cryptocurrency in order to utilize the malevolent chatbot.

The likelihood of these websites being fraudulent is heightened by the fact that the creator of WormGPT purportedly abandoned the project after their identity was uncovered last month. As reported by security journalist Brian Krebs, the individual behind WormGPT is a 23-year-old Portuguese programmer named Rafael Morais, who has since distanced himself from promoting his chatbot for malicious purposes.

Subsequent to this revelation, the user account endorsing WormGPT announced on a hacking forum that their team was discontinuing the project. 

The account expressed, "With great sadness, I come to inform everyone about the end of the WormGPT project. From the beginning, we never thought we would gain this level of visibility, and our intention was never to create something of this magnitude."

In the weeks leading up to the shutdown, the official WormGPT account on Telegram also issued a warning about scammers imitating the chatbot’s brand. The post stated, "We don’t have any website and either any other groups in any platform. The rest are resellers or scammers!"

Later, the same account expressed disbelief at how people were still falling victim to scams in 2023.

Despite the apparent shutdown of WormGPT, the menace of cybercriminals turning to other malevolent chatbots for aid remains. In July, a different developer was observed creating FraudGPT.