A Critical Security Issue Has Been Detected in Google’s Artificial Intelligence Gemini!

There has been a remarkable development regarding Gemini, the productive artificial intelligence model of the US-based technology giant Google. It was revealed that Google’s productive artificial intelligence Gemini scanned PDF files in Google Drive without user permission. While new concerns about the security of artificial intelligence emerged with the situation, how it was detected became a matter of curiosity. Here are the details…

Google Gemini Doesn’t Need Permission When Reading PDFs

While reviewing his tax return in Google Docs, AI consultant Kevin Bankston noticed that Gemini was automatically summarizing the return. Bankston tried looking for settings to prevent Gemini from reading PDFs in Drive, but couldn’t find any. Bankston tried to report the situation to Gemini but received no response.

Bankston argues that this situation could lead to major security vulnerabilities. However, in his investigations, he also states that this situation only occurs if Gemini is activated in a file.

Google did not make any statement on the subject. It’s obvious that users won’t want Gemini to read their PDFs. Google needs to fix this issue and provide more comprehensive settings options for Gemini.

This incident draws attention to the increasing power of artificial intelligence and the ethical problems it brings. It is questioned how transparent and user-controlled artificial intelligence should be.