Dynamically redacting confidential information

1. A computer-implemented method for dynamically redacting confidential audio information from communications, the method comprising:

receiving, by a communications analyzer, an encrypted audio communication and identification of an intended recipient of the audio communication from a security application;

decrypting the encrypted audio communication;

identifying, by the communications analyzer, tokens within the audio communication;

comparing, by the communications analyzer, the identified tokens to classified tokens in a classified token database and determining whether any of the identified tokens match classified tokens in the classified token database, wherein the comparison includes using natural language processing and/or audio analysis to compare the identified tokens to the classified tokens in the classified token database;

receiving, by the communications analyzer from the token database, a list of privileged recipients derived from a user access list of privileges and/or classifications associated with users;

determining, by the communications analyzer, whether the intended recipient is on the list of privileged recipients;

determining, by the communications analyzer, whether to perform a redaction on the audio communication based at least in part on the comparison and a determination that the intended recipient is not on the list of privileged recipients;

in response to determining that the redaction on the audio communication is to be performed and in response to finding at least one identified token that matches a classified token in the classified token database, selecting, by the communications analyzer, a portion of the audio communication corresponding to the at least one identified token to redact, wherein the audio communication is intended for multiple recipients, wherein the redaction is to be performed for the recipient that is not on the list of privileged recipients; and

outputting, by the communications analyzer, to the security application, an indication of the selected portion of the audio communication to enable redaction of the selected portion by a remote user device of a sender of the audio communication,

wherein the indication also includes an option, to the sender of the audio communication, of whether to perform the redaction.

2. The computer-implemented method of claim 1 , comprising sending the indication of the selected portion to a remote user device to enable the user device to redact the selected portion.

3. The computer-implemented method of claim 1 , comprising sending a recommendation to send the unredacted audio communication to a second recipient that is on the list of privileged recipients using a particular communication medium which is relatively more secure than the communication medium with which the redacted audio communication is to be sent to the recipient that is not on the list of privileged recipients, wherein the relatively more secure communication medium is selected from the group consisting of: a secure company email account, a security application, and a preauthorized application.

4. The computer-implemented method of claim 1 , wherein at least some information in the classified token database is indicative of a non-privileged recipient, wherein the at least some information is selected from the group consisting of: a name of a non-privileged recipient, a title of a non-privileged recipient, and a server which hosts a communication medium which is not secure.

5. The computer-implemented method of claim 1 , wherein at least some information in the classified token database is indicative of a communication medium which is not secure, wherein the at least some information is selected from the group consisting of: a name of a communication medium and a server which hosts a communication medium which is not secure; wherein the communication medium which is not secure is selected from the group consisting of: a personal email account, a text message, and a social media application.

6. The computer-implemented method of claim 1 , wherein at least some of the classified tokens in the classified token database are indicative of information which is not publicly available, wherein the information which is not publicly available is selected from the group consisting of: a product name, a product release date, design documentation, and intellectual property documentation.

7. The computer-implemented method of claim 1 , wherein selecting the portion of the audio communication corresponding to the at least one identified classified token for redaction includes selecting a portion from the group consisting of: a full sentence containing the at least one identified classified token, a full paragraph containing the at least one identified classified token, an entire attachment containing the at least one identified classified token, and the entire audio communication containing the at least one identified classified token.

8. The computer-implemented method of claim 1 , wherein at least some of the classified tokens are associated with a subject selected from the group consisting of: a project, a business unit, and a team.

9. The computer-implemented method of claim 1 , wherein the determination of whether to perform the redaction on the audio communication is based at least in part on a factor selected from the group consisting of: whether a communication medium by which the audio communication is to be sent is not secure, and whether a recipient is not a privileged recipient.

10. The computer-implemented method of claim 1 , in response to determining that the redaction on the audio communication is to be performed, send an alert to a user and/or at least one recipient of the audio communication.

11. The computer-implemented method of claim 1 , wherein the audio communication is sent to more than one recipient and the redaction is applied to at least a portion of the audio communication to at least one recipient.

12. The computer-implemented method of claim 1 , wherein the natural language processing parses the audio communication into confidentiality levels.

13. The computer-implemented method of claim 12 , wherein the confidentiality levels correspond to at least some of the identified tokens in the audio communication.

14. The computer-implemented method of claim 12 , wherein the confidentiality levels correspond to at least one recipient.

15. A computer program product for dynamically redacting confidential information from audio communications, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to:

receive an encrypted audio communication and identification of an intended recipient of the audio communication from a security application;

decrypt the encrypted audio communication;

identify tokens within the audio communication;

compare the identified tokens to classified tokens in a classified token database and determining whether any of the identified tokens match classified tokens in the classified token database, wherein the comparison includes using natural language processing to compare the identified tokens to the classified tokens in the classified token database;

determine, from the token database, a list of privileged recipients derived from a user access list of privileges and/or classifications associated with users;

determine whether the intended recipient is on the list of privileged recipients;

determine whether to perform a redaction on the audio communication based at least in part on the comparison and a determination that the intended recipient is not on the list of privileged recipients;

in response to determining that the redaction on the audio communication is to be performed and in response to finding at least one identified token that matches a classified token in the classified token database, select a portion of the audio communication corresponding to the at least one identified token to redact, wherein the audio communication is intended for multiple recipients, wherein the redaction is to be performed for the recipient that is not on the list of privileged recipients; and

output, to the security application, an indication of the selected portion of the audio communication to enable redaction of the selected portion by a remote user device of a sender of the audio communication,

wherein the indication also includes an option, to the sender of the audio communication, of whether to perform the redaction.

16. The computer program product of claim 15 , wherein the natural language processing parses the audio communication into confidentiality levels.

17. The computer program product of claim 16 , wherein the confidentiality levels correspond to at least some of the identified tokens in the audio communication, wherein the confidentiality levels correspond to at least one recipient.

18. A system, comprising:

a hardware processor; and

logic integrated with the processor, executable by the processor, or integrated with and executable by the processor, the logic being configured to:

receive an encrypted audio communication and identification of an intended recipient of the audio communication from a security application;

decrypt the encrypted audio communication;

identify tokens within the audio communication;

compare the identified tokens to classified tokens in a classified token database and determining whether any of the identified tokens match classified tokens in the classified token database, wherein the comparison includes using natural language processing to compare the identified tokens to the classified tokens in the classified token database;

receive, from the token database, a list of privileged recipients derived from a user access list of privileges and/or classifications associated with users;

determine whether the intended recipient is on the list of privileged recipients;

determine whether to perform a redaction on the audio communication based at least in part on the comparison and a determination that the intended recipient is not on the list of privileged recipients;

in response to determining that the redaction on the audio communication is to be performed and in response to finding at least one identified token that matches a classified token in the classified token database, select a portion of the audio communication corresponding to the at least one identified token to redact, wherein the audio communication is intended for multiple recipients, wherein the redaction is to be performed for the recipient that is not on the list of privileged recipients; and

output, to the security application, an indication of the selected portion of the audio communication to enable redaction of the selected portion by a remote user device of a sender of the audio communication,

wherein the indication also includes an option, to the sender of the audio communication, of whether to perform the redaction.

19. The system of claim 18 , wherein the natural language processing parses the audio communication into confidentiality levels.

20. The system of claim 19 , wherein the confidentiality levels correspond to at least some of the identified tokens in the audio communication, wherein the confidentiality levels correspond to at least one recipient.