Google to Use Hashes to Remove Nonconsensual Images from Search

Google is stepping up its fight against harmful content. Starting soon, Google will use hashes to find and remove nonconsensual intimate imagery from Search, working with StopNCII.org to protect victims from exploitation. This move aims to make the web safer while reducing the burden survivors face in reporting and removing abusive content.

How the system works

Hashes are unique digital fingerprints created from images or videos. By using StopNCII’s technology, Google can detect nonconsensual intimate imagery without storing or sharing the original files. For images, the system uses PDQ hashing, while videos rely on MD5. This ensures privacy is maintained while harmful content is removed from search results.

Why Google is adopting this now

While Facebook, Instagram, TikTok, Bumble, and Microsoft’s Bing adopted StopNCII years ago, Google has been slower to roll out this safeguard. Critics have pushed the company to do more, and Google’s own blog post admits there’s “more to be done” to support victims at scale. With this update, Google is finally aligning with industry peers in tackling the problem.

What this means for victims

Previously, survivors had to identify and flag abusive content themselves, often reliving trauma in the process. Now, by leveraging StopNCII’s global database of hashes, Google can proactively block nonconsensual images and revenge porn. This is especially important with the rise of AI-generated intimate imagery, which is increasingly difficult for victims to track and report.

A step forward in online safety

While no system is perfect, this change represents progress in reducing the spread of abusive content online. Google’s use of hashes may not eliminate all harmful imagery, but it shifts responsibility away from victims and toward a more proactive model of protection.