According to AIbase, Google Chrome recently introduced a key update that allows users to independently disable and remove the AI model stored locally on their devices for fraud detection. This change marks Google's effort to promote AI security features while providing users with more detailed system resource management options.
Last year, Chrome upgraded the underlying technology of its "Enhanced Protection" feature, introducing an on-device AI model to provide "real-time" protection. This model can identify and block dangerous websites, malicious downloads, and suspicious extensions that are not yet recorded in Google's database, and perform deep scans on high-risk files.
According to the latest findings from security researcher Leo, users can now fully control the presence or absence of this model. If users wish to remove the model due to system performance or privacy concerns, they can simply go to Chrome Settings > System menu and turn off the "On-device GenAI" switch.
This management switch is currently available in Chrome Canary (Canary version) and is expected to be rolled out to all stable version users soon. Notably, this on-device AI infrastructure will not only serve fraud detection but also support more generative AI features in Chrome in the future.
