Security flaws discovered in ChatGPT systems and Mac app

Summarised from 2 outlets · Updated 2 Oct, 11:31 · Archive

Hackers hosted malware disguised as a fake ChatGPT model, while a vulnerability in the Mac app could have exposed sensitive data.

Two security flaws have been discovered in ChatGPT systems. TechRadar reports that hackers hosted a fake ChatGPT model on what appeared to be the official website, which actually distributed malware rather than providing the expected AI service. The fake model included a fake CAPTCHA as part of the attack.

Separately, Wired reports that a vulnerability in ChatGPT's Mac application could have permitted hackers to extract sensitive data from users. The flaw has now been patched. Both incidents highlight that AI software itself remains an attractive target for cyber attacks despite growing focus on the hacking capabilities of AI systems themselves.

How it is being reported

• A Flaw in ChatGPT’s Mac App Could Have Let Hackers Grab Sensitive DataWired · 2 Oct, 10:45
• Hackers host fake ChatGPT model on its official website — but really it's just malwareTechRadar · 30 Sept, 17:10

In this story: ChatGPT · TechRadar · Wired

This summary was written by AI from the headlines and standfirsts above, and states as fact only what at least two outlets report. How we use AI · Report a problem