In the context of a computer client-server architecture, typically used in the Internet for communicating between a server and applications running on user computers (clients), a method is provided for enhancing security in the context of digital rights management (DRM) where the server is an untrusted server that may not be secure, but the client is secure. This method operates to authenticate the server to the client and vice versa to defeat hacking attacks intended to obtain confidential information. Values passed between the server and the client include encrypted random numbers, authentication values and other verification data generated using cryptographic techniques including double encryption.