Cryptographic systems often rely on the secrecy of cryptographic credentials; however, these are vulnerable to eavesdropping and can resist neither a user's intentional disclosure nor coercion attacks where the user is forced to reveal the credentials. Conventional biometric keys (e.g., fingerprint, iris, etc.), unfortunately, can still be surreptitiously duplicated or adversely revealed. In this research, the PIs argue that the most secure cryptographic credentials are ones of which the users aren't even aware.