Digital Identity Theft Using Deepfakes
Alisha Gilbert, Zhigang Gong · 2024
This work aims to discuss aspects of identity theft by deepfakes. Deepfakes are convincing videos digitally manipulated to present people’s speeches and actions that have never happened. A person’s behavior, facial expression, and voice can be mimicked by neural networks on which deepfakes depend. Deepfakes could become a threat to our identities. Cybercriminals can use deepfakes to steal our identities so that they can access or create online accounts and commit fraud in our names. A theoretical framework for addressing the identity theft by deepfakes problem comprises procedures of legislation enforcement, security policy revision, and security mechanism enhancement. Governments, such as the US and China, have enacted new laws that criminalize certain deepfakes. Security policies should be revised to add provisions on how to deal with requests or orders originating from phone calls and voice chats. Employees should be trained to realize that sharing lots of personal photos and voice snippets on social media makes them vulnerable to deepfakes. Appropriate access control and security mechanisms should be enhanced in organizations. Authentication services that use biometric facial recognition should be tested to verify that they can detect deepfakes.