Phishing attempts using artificially synthesized audio (“deep fakes”) are increasingly used to deceive users that receive voice calls via telephone networks or calling apps. Voice cloning technology can be used to conduct a fake phone call in which the caller audio is faked to imitate someone else’s identity.

The techniques described in this disclosure analyze an incoming call and assist a user in the verification of caller identity. Fake voice calls are detected automatically and alerts are provided to the user. The detection is based on user-permitted information such as caller phone number, profile information, voice profiles of callers, and other contextual factors. Fake calls can be detected prior to the user answering the call or during an ongoing call. The described techniques are implemented on-device and with specific user permission.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.