A familiar cloud of doubt is drifting back over Google… and it’s hovering above the photo memories of over a billion people.
According to a report by Forbes, Google is facing renewed scrutiny after a public accusation that images stored in Google Photos could be feeding its powerful AI systems. The claim did not come from regulators or whistleblowers, but from Proton, a privacy-focused tech company and a direct competitor in cloud storage.
The timing raised eyebrows: The allegation surfaced shortly after Google’s new AI image tool, Nano Banana, drew widespread attention for producing images so realistic that critics say the line between real and artificial is quickly fading.
The controversy began with a post from Proton on X that quickly spread across social media.
“When you know the only reason Google’s AI is the best at generating images because they’re scanning every Android user’s Google Photos albums but they won’t admit it & you can’t prove it,” Proton wrote.
According to Forbes, the statement offered no evidence. But it was enough to ignite fears among users already wary of how much data big tech companies collect and analyze.
Google responds to the accusation
Google has strongly denied the accusation.
In a statement shared with Forbes, the company said, “we don’t train any generative AI models outside of Photos with your personal data, which includes other Gemini models and products. If you choose to share a photo or video from Google Photos with other Google or third-party services, your data is processed according to those other services’ policies.”
While Google rejects the AI training claim, it does acknowledge that Google Photos is not an end-to-end encrypted service. Images stored on the platform are scanned for child sexual abuse material using automated systems and trained specialist teams, a practice Google says is aimed at safety and compliance.
The controversy highlights a larger problem facing cloud services. AI tools are becoming deeply embedded in everyday platforms, often governed by long and complex privacy policies that many users never read. When claims like this surface, they can quickly undermine trust, even without proof.
As Zak Doffman writes in Forbes, the takeaway for users is familiar but important: cloud platforms are not fully private unless they use end-to-end encryption. Photos stored online may be analyzed in some form, even if companies say they are not being used for AI training.
Want more control over your Google account? Here’s how Google now lets users change their Gmail address without creating a new account.
