Google Chrome now allows users to delete the local AI models that power its “Enhanced Protection” security feature, giving them greater control over on-device artificial intelligence. The option was spotted in Chrome Canary by browser enthusiast Leo (@Leopeva64).
Google upgraded Enhanced Protection last year by integrating AI models to offer “real-time” defence against dangerous websites, downloads and extensions. While the company has not fully explained how the new version differs from the older one, it confirmed that the feature uses AI to analyse patterns in real time, helping Chrome warn users about threats such as phishing sites or suspicious downloads even before Google officially identifies them.
With the latest update, users can now manage these models directly. By going to Chrome Settings > System and turning off “On-device GenAI,” the local AI model is removed from the device.
Reports suggest the local model’s role could expand beyond scam detection to support other Chrome features in the future. The setting is currently available in the experimental Chrome Canary build, and Google is expected to roll it out to all users soon.
