Threat detection with business impact scoring

1. A computer program product comprising computer executable code embodied in a non-transitory computer readable medium that, when executing on one or more computing devices, performs the steps of:

providing a valuation model for automatically estimating a business value of a file, wherein providing the valuation model includes training a machine learning model to estimate the business value based on a training set of files each having a known business value, wherein the business value is based on at least one of file location or an access control list;

creating an integrative model that evaluates a potential threat by a threat sample based on a combination of a first model configured to identify malicious code based on behavioral tags, a second model configured to identify malicious code based on an executable file path, and a third model configured to identify malicious code based on a Uniform Resource Locator within the threat sample;

configuring a threat management facility to evaluate new threat samples on endpoints within an enterprise network according to the integrative model;

identifying one or more intermediate threats by any of the new threat samples that are not within a predetermined confidence level of safe code or malicious code according to the integrative model;

providing a user interface for presenting the one or more intermediate threats to a user for human evaluation; and

ranking the one or more intermediate threats for presentation within the user interface based on a combination of a likelihood of maliciousness determined according to the integrative model and an estimated business value, determined according to the valuation model, of files associated with the intermediate threat.

2. The computer program product of claim 1 wherein the business value is further based on one or more of file content, encryption status, file type, file usage history, file creation date, file modification date, or file author.

3. A method comprising:

identifying, in a threat sample, one or more intermediate threats to an enterprise network with an integrative model, wherein the integrative model is based on a combination of a first model configured to identify malicious code based on behavioral tags, a second model configured to identify malicious code based on an executable file path, and a third model configured to identify malicious code based on a Uniform Resource Locator within the threat sample, wherein the one or more intermediate threats include one or more computing objects with an objective score from the integrative model that are not within a predetermined confidence level of a safe score or a malicious score;

automatically estimating a business value associated with each of the one or more intermediate threats with a valuation model, wherein the valuation model includes a machine learning model trained to estimate the business value based on a training set of files each having a known business value, wherein the business value is based on at least one of file location or an access control list; and

presenting a list of the one or more intermediate threats in a user interface for human evaluation, the list ranked according to a combination of the objective score and the business value.

4. The method of claim 3 wherein the the business value is further based on one or more of file content, encryption status, file type, file usage history, file creation date, file modification date, or file author.

5. The method of claim 3 further comprising receiving a user-initiated remedial action for one of the intermediate threats in the user interface.

6. The method of claim 3 further comprising receiving a user risk assessment for one of the intermediate threats in the user interface.

7. The method of claim 3 further comprising remediating a risk to a high business value computing object in response to a user input in the user interface.

8. The method of claim 3 wherein the integrative model evaluates a potential threat by computer objects based on one or more of file behavior, file signature, file path, or Uniform Resource Locator.

9. The method of claim 3 wherein the integrative model includes one or more machine learning models trained to recognize potentially malicious code based on a training set of known safe and known unsafe threat samples.

10. The method of claim 3 wherein the one or more computing objects include at least one of a process, an executable, or a file.

11. The method of claim 3 wherein the one or more computing objects include at least one of a registry of system settings or a secure kernel cache of process information.

12. A system comprising:

a memory storing an integrative model and a valuation model, the integrative model configured to evaluate a potential threat by a threat sample based on a combination of a first model configured to identify malicious code based on behavioral tags, a second model configured to identify malicious code based on an executable file path, and a third model configured to identify malicious code based on a Uniform Resource Locator within the threat sample, and the valuation model configured to automatically estimate a business value of one or more files associated with the threat sample, wherein the valuation model includes a machine learning model trained to estimate the business value based on a training set of files each having a known business value, wherein the business value is based on at least one of file location or an access control list;

a threat management facility configured to apply the integrative model to new threat samples and to identify intermediate threats that are not within a predetermined likelihood of being safe or unsafe; and

a web server configured to display a list of intermediate threats in a user interface for human evaluation, wherein the list of intermediate threats is ranked according to a combination of a first score from the integrative model and a second score from the valuation model.

13. The system of claim 12 wherein the threat management facility is configured to remediate a risk to an endpoint in response to a user input received through the user interface.