Chrome Quietly Removed Its Privacy Promise for On-Device AI — What You Need to Know

If you use Chrome and have started noticing a new AI feature—like suggested replies or automated text rewriting—you might have assumed it works entirely on your device. Google previously said exactly that. But recently, without any announcement, the company deleted that privacy promise from its support pages. This article explains what changed, why it matters, and how you can protect yourself.

What happened

Google had a line in Chrome’s privacy FAQ that read: “On-device AI features process data locally and do not send information to Google servers.” That statement is no longer there. Multiple outlets, including Decrypt, Yahoo Tech, and GIGAZINE, confirmed the removal between late April and early May 2026.

At the same time, Chrome has been silently downloading a 4 GB AI model onto users’ computers. Even if you delete that model manually, Chrome can re-download it on the next update or restart. The feature appears to be part of Google’s push to integrate generative AI—like “Help me write” and smart reply—directly into the browser.

Google has not issued a public statement explaining why the privacy promise was removed or whether data now may be sent to its servers.

Why it matters

For years, “on-device” was a specific guarantee: the processing happens locally, no data leaves your machine. That assurance was important for anyone using Chrome for sensitive tasks—drafting private emails, composing messages, or analyzing images.

Removing that language without explanation creates uncertainty. You can no longer be sure that an AI feature labeled “on-device” is truly offline. If Google is now allowing some data to be uploaded, it could affect your privacy in ways you didn’t consent to. The silent nature of the change is especially concerning because users were never told their privacy expectation was being altered.

What readers can do

You don’t have to accept this change. Here are practical steps to regain control:

  1. Check your Chrome AI settings.
    Type chrome://settings/ai into the address bar (or go to Settings → Privacy and security → AI). You’ll see a list of features like “Help me write,” “Tab organizer,” or “Smart reply.” Look for any option that says “Use on-device AI” or similar.

  2. Disable all AI features.
    Turn off every switch in that settings page. This prevents Chrome from downloading or running any AI model. Some features may need a browser restart to fully disengage.

  3. Delete the downloaded model (if present).
    In the same settings area, you might see an option to clear the downloaded AI model. Alternatively, you can find it on disk—often in Chrome’s user data folder. However, note that Chrome may re-download it later. Keeping AI features disabled is the more reliable method.

  4. Consider using a different browser.
    If you’re uncomfortable with Google’s approach, Firefox and Brave offer similar functionality (if you want it) with clearer privacy policies. Both are open-source and don’t silently install large AI models. For everyday browsing, they work just as well as Chrome.

  5. Keep an eye on Chrome’s privacy FAQ.
    Bookmark the Chrome Privacy page and check it periodically. If Google changes its stance again, you’ll see it—but don’t rely on that alone.

Sources

  • Decrypt: “Chrome Deleted Its Own Privacy Promise for Sneaky On-Device AI” (May 7, 2026)
  • Yahoo Tech: “Chrome Deleted Its Own Privacy Promise for Sneaky On-Device AI” (May 7, 2026)
  • GIGAZINE: “Chrome removes claim that its ‘on-device AI’ does not send data to Google servers” (May 9, 2026)

None of these reports have received an official response from Google as of this writing.