Secure voice interface in a streaming media device to avoid vulnerability attacks

A method for receiving secure voice commands by an application executing within a smart media device while attached to the Internet, the Internet communicatively interconnecting the smart media device with a remote voice-speech server and one or more content servers, the smart media device having a secure microphone input for accepting audio data containing voice commands and providing encrypted audio data using a pre-stored encryption key and a trusted execution environment for decrypting the encrypted audio data from the secure microphone input, the method comprising:

• based on determining that the application uses secure voice commands and the remote voice-speech server, performing the following:
• enabling the trusted execution environment to receive and decrypt the encrypted audio data;
• sending the decrypted audio data to the remote voice-speech server for processing so as to cause the remote voice-speech server to generate one or more application commands;
• receiving, by the application, the one or more application commands;
• and
• performing, by the application, the one or more application commands;
• and
• based on determining that the secure voice commands are not to be processed by the remote voice-speech server, perform the following:
• receiving the encrypted audio data in the trusted execution environment;
• receiving and decrypt, within the trusted execution environment, the encrypted audio data;
• generating, within the trusted environment, one or more other application commands corresponding to processing decrypted voice commands;
• receiving, by the application, the one or more other application commands;
• and
• performing, by the application, the one or more other application commands.

A non-transitory computer-readable recording medium in a smart media for receiving secure voice commands by an application executing within the smart media device while attached to the Internet, the Internet communicatively interconnects the smart media device with a remote voice-speech server and one or more content servers, the smart media device having a secure microphone input for accepting audio data containing voice commands and providing encrypted audio data using a pre-stored encryption key and a trusted execution environment for decrypting the encrypted audio data from the secure microphone input, the non-transitory computer-readable recording medium storing one or more programs which when executed by a controller of the smart media device performs a method comprising:

• based on determining that the application uses secure voice commands and the remote voice-speech server, performing the following:
• enabling the trusted execution environment to receive and decrypt the encrypted audio data;
• sending the decrypted audio data to the remote voice-speech server for processing so as to cause the remote voice-speech server to generate one or more application commands;
• receiving, by the application, the one or more application commands;
• and
• performing, by the application, the one or more application commands;
• and
• in response to determining that the secure voice commands are not to be processed by the remote voice-speech server, performing the following:
• enabling the trusted execution environment to receive the encrypted audio data;
• receiving and decrypting, by the trusted execution environment, the encrypted audio data;
• generating the one or more other application commands corresponding to processing decrypted voice commands, by the trusted execution environment;
• receiving, by the application, the one or more other application commands;
• and
• performing, by the application, the one or more other application commands.