Gemini AI caught scanning Google Drive hosted PDF files without permission — user complains feature can’t be disabled

As part of the wider tech industry’s wider push for AI, whether we want it or not, it seems that Google’s Gemini AI service is now reading private Drive documents without express user permission, per a report from Kevin Bankster on Twitter embedded below. While Bankster goes on to discuss reasons why this may be glitched for users like him in particular, the utter lack of control being given over his sensitive, private information is unacceptable for a company of Google’s stature —and does not bode well for future privacy concerns amongst AI’s often-forced rollout.

So, what exactly is going on here? Both Google support and the Gemini AI itself do not quite seem to know, but Kevin Bankston has some theories, after providing much more detail in the full thread. Contrary to the initial posting, this is technically happening within the larger umbrella of Google Drive and not Google Docs specifically, though it seems likely the issue could apply to Docs as well.

But what caused this issue? According to Google’s Gemini AI, the privacy settings used to inform Gemini should be openly available, but they aren’t, which means the AI is either “hallucinating (lying)” or some internal systems on Google’s servers are outright malfunctioning. Either way, not a great look, even if this private data supposedly isn’t used to train the Gemini AI.

What’s more, Bankston did eventually find the settings toggle in question… only to find that Gemini summaries in Gmail, Drive, and Docs were already disabled. Additionally, it was in an entirely different place than either of the web pages to which Gemini’s bot initially pointed.

For Bankston, the issue seems localized to Google Drive, and only happens after pressing the Gemini button on at least one document. The matching document type (in this case, PDF) will subsequently automatically trigger Google Gemini for all future files of the same type opened within Google Drive. He additionally theorizes that it may have been caused by him enabling Google Workspace Labs back in 2023, which could be overriding the intended Gemini AI settings.

Even if this issue is isolated to Google Workspace Labs users, it’s quite a severe downside for having helped Google test its latest and greatest tech. User consent still matters on a granular basis, particularly with potentially sensitive information, and Google has utterly failed at least one segment of its user base by failing to stay true to that principle.