Detecting endpoint compromise based on network usage history

1. A computer program product for protecting against malicious network activity in an enterprise network, the computer program product comprising computer executable code embodied in a non-transitory computer readable medium that, when executing on a processor of an endpoint operated by an end user in the enterprise network, performs the steps of:

detecting a connection initiated from an application executing on the endpoint in the enterprise network to a network address outside the enterprise network;

receiving a reputation of the application that is based on a first reputation lookup at a threat management facility for the enterprise network;

receiving a reputation of the network address that is based on a second reputation lookup at the threat management facility;

determining a network usage history for the application using a log of network activity maintained on the endpoint, the network usage history providing a general pattern of usage associated with the application on the endpoint, the general pattern of usage including a list of URLs associated with historic usage for the application on the endpoint;

evaluating the endpoint for a compromised condition based on the reputation of the application, the reputation of the network address, and the network usage history for the application, including at least a determination by the processor of whether the network address in the connection initiated from the application is outside historic usage for the application on the endpoint based on whether the network address is included on the list of URLs for resources outside the enterprise network that are associated with the historic usage for the application on the endpoint; and

initiating remediation action on the endpoint when the compromised condition is detected.

2. The computer program product of claim 1 , wherein the reputation of the application is at least one of known malicious, suspect, unknown, or known good.

3. The computer program product of claim 1 , wherein the reputation of the network address is at least one of known malicious, suspect, unknown, or known good.

4. The computer program product of claim 1 , wherein the reputation of the network address is based on crowd-sourced information about the network address.

5. The computer program product of claim 1 , further comprising code that performs the step of monitoring network activity by the application and storing the network usage history for the application in the log of network activity maintained on the endpoint.

6. The computer program product of claim 1 , further comprising code that performs the step of associating a network communication to the network address by a service executing on the endpoint with the application when the application controls the network communication, and adding the network communication to the network usage history for the application.

7. A method performed by computing circuitry of an endpoint user device, the method comprising:

detecting a connection initiated from an application executing on an endpoint in an enterprise network to a network address;

receiving, by using the computing circuitry of the endpoint user device, a reputation of the application using the connection, wherein the reputation of the application is requested from a threat management facility for the enterprise network;

receiving a reputation of the network address, wherein the reputation of the network address is requested from a threat management facility for the enterprise network;

determining a network usage history for the application using a log of network activity maintained on the endpoint user device, the network usage history providing a general pattern of usage associated with the application on the endpoint, the general pattern of usage including a list of URLs associated with historic usage for the application on the endpoint;

evaluating the endpoint for a compromised condition based on the reputation of the application, the reputation of the network address, and the network usage history for the application, including at least a determination by the computing circuitry of whether the network address in the connection initiated from the application is outside historic usage for the application on the endpoint based on whether the network address is included on the list of URLs for resources outside the enterprise network that are associated with the historic usage for the application on the endpoint; and

initiating remediation action on the endpoint when the compromised condition is detected.

8. The method of claim 7 , wherein the reputation of the application is at least one of known malicious, suspect, unknown, or known good.

9. The method of claim 7 , wherein the reputation of the network address is at least one of known malicious, suspect, unknown, or known good.

10. The method of claim 7 , wherein the reputation of the network address is based on crowd-sourced information about the network address.

11. The method of claim 7 , further comprising monitoring network activity by the application and storing the network usage history for the application in a log on the endpoint.

12. The method of claim 11 , further comprising associating a network communication to the network address by a service executing on the endpoint with the application when the application controls the network communication, and adding the network communication to the network usage history for the application.

13. A threat management system comprising:

a network interface for coupling the threat management system to an enterprise network;

a memory;

a processor configured by computer executable code stored in the memory to protect against malicious network activity in the enterprise network by performing the steps of:

detecting a connection initiated from an application executing on an end user endpoint in the enterprise network to a network address,

determining a reputation of the application using the connection,

determining a reputation of the network address; and

a local security agent of the end user endpoint configured by computer executable code stored in a second memory to perform the steps of:

determining a network usage history for the application using a log of network activity maintained on the end user endpoint, the network usage history providing a general pattern of usage associated with the application on the end user endpoint, the general pattern of usage including a list of URLs associated with historic usage for the application on the end user endpoint,

evaluating the end user endpoint for a compromised condition based on the reputation of the application, the reputation of the network address, and the network usage history for the application, including at least a determination of whether the network address in the connection initiated from the application is outside historic usage for the application on the end user endpoint based on whether the network address is included on the list of URLs for resources outside the enterprise network that are associated with the historic usage for the application on the end user endpoint, and

initiating remediation action on the end user endpoint when the compromised condition is detected.

14. The threat management system of claim 13 , wherein the reputation of the application is at least one of known malicious, suspect, unknown, or known good and wherein the reputation of the network address is at least one of known malicious, suspect, unknown, or known good.