Cybercriminals cracking ChatGPT to enable cybercrime for the masses

Cybercriminals have shown sufficient prowess in hacking and harnessing the capabilities of the artificial intelligence (AI) chatbot ChatGPT to warrant an early warning to cyber clients on emerging threats.

“There's evidence of nefarious activity beginning to surface,” cyber leaders at global insurance broker AJ Gallagher said in a post to clients. “We're starting to see evidence of malicious actors plotting attacks via this new threat vector.”

Early signs point to success in malware development and the crafting of phishing emails, Gallagher's managing director of its cyber liability practice John Farley wrote, citing discoveries from cyber security firms.

Underground forums have been ablaze with talk of workarounds, often based on Telegram communicator and the platform's own programming interface, to the barriers to malicious content written in by ChatGPT’s developer, OpenAI, one of those studies claims.

That could open the jacking industry to a flood of novice practitioners, willing to pay higher-end cybercriminals for the right to craft targeted malware and phishing endeavours, Gallagher warns.

“As sophisticated hackers find ways to exploit ChatGPT and sell their services on a mass scale, we may see an exponentially greater number of hackers emerge this year and into the future,” Farley wrote.

“At this early stage their efforts appear crude at best, but we do foresee a time in the near future where ChatGPT-related exploits may become both sophisticated and widespread.”

Did you get value from this story?  Sign up to our free daily newsletters and get stories like this sent straight to your inbox.