The most recent generative AI fashions aren’t simply stand-alone text-generating chatbots—as a substitute, they’ll simply be hooked as much as your knowledge to offer customized solutions to your questions. OpenAI’s ChatGPT can be linked to your Gmail inbox, allowed to examine your GitHub code, or discover appointments in your Microsoft calendar. However these connections have the potential to be abused—and researchers have proven it might probably take only a single “poisoned” doc to take action.
New findings from safety researchers Michael Bargury and Tamir Ishay Sharbat, revealed on the Black Hat hacker convention in Las Vegas as we speak, present how a weak point in OpenAI’s Connectors allowed delicate info to be extracted from a Google Drive account utilizing an indirect prompt injection attack. In an indication of the assault, dubbed AgentFlayer, Bargury exhibits the way it was doable to extract developer secrets and techniques, within the type of API keys, that had been saved in an indication Drive account.
The vulnerability highlights how connecting AI fashions to exterior methods and sharing extra knowledge throughout them will increase the potential assault floor for malicious hackers and doubtlessly multiplies the methods the place vulnerabilities could also be launched.
“There may be nothing the person must do to be compromised, and there’s nothing the person must do for the info to exit,” Bargury, the CTO at safety agency Zenity, tells WIRED. “We’ve proven that is utterly zero-click; we simply want your e mail, we share the doc with you, and that’s it. So sure, that is very, very dangerous,” Bargury says.
OpenAI didn’t instantly reply to WIRED’s request for remark concerning the vulnerability in Connectors. The corporate launched Connectors for ChatGPT as a beta characteristic earlier this 12 months, and its website lists not less than 17 totally different companies that may be linked up with its accounts. It says the system means that you can “convey your instruments and knowledge into ChatGPT” and “search information, pull stay knowledge, and reference content material proper within the chat.”
Bargury says he reported the findings to OpenAI earlier this 12 months and that the corporate rapidly launched mitigations to stop the method he used to extract knowledge through Connectors. The best way the assault works means solely a restricted quantity of knowledge may very well be extracted directly—full paperwork couldn’t be eliminated as a part of the assault.
“Whereas this problem isn’t particular to Google, it illustrates why creating strong protections in opposition to immediate injection assaults is vital,” says Andy Wen, senior director of safety product administration at Google Workspace, pointing to the corporate’s recently enhanced AI security measures.
