Recommendation interface for remediation plans for cyberattack vulnerabilities
Systems and methods are provided to build a machine learned exploitability risk model that predicts, based on the characteristics of a set of machines, a normalized risk score quantifying the risk that the machines are exploitable by a set of attacks. To build the model, a training dataset is constructed by labeling characteristic data of a population of machines with exploitation test results obtained by simulating a set of attacks on the population. The model is trained using the training data to accurately predict a probability that a given set of machines is exploitable by the set of attacks. In embodiments, the model may be used to make quick assessments about how vulnerable a set of machines are to the set of attacks. In embodiments, the model may be used to compare the effectiveness of different remediation actions to protect against the set of attacks.
1. A system, comprising:
one or more hardware processors with associated memory that implement a remediation planning system, configured to:
determine, using an exploitability risk model trained according to a machine learning technique, a risk score of a set of machines for an attack based on characteristics data of the set of machines;
determine a set of candidate remediation plans to reduce the risk score;
apply the candidate remediation plans to the characteristics data to generate updated instances of the characteristics data, wherein the application simulates performance of the candidate remediation plans on the set of machines without performing the candidate remediation plans on the set of machines;
determine, using the exploitability risk model and based on the updated instances of characteristics data, updated risk scores of the set of machines after the simulated performance of respective ones of the candidate remediation plans;
generate output via a graphical user interface (GUI) of the remediation planning system, wherein the GUI indicates, the attack, the risk score, the candidate remediation plans, the updated risk scores, and a ranking of the candidate remediation plans determined based at least in part on the updated risks scores; and
responsive to user selection of one of the candidate remediation plans received via the GUI, generate a workflow to perform the one candidate remediation plan on the set of machines.
2. The system of claim 1 , wherein the remediation planning system is configured to:
select a plurality of remediation actions from a remediation action library based on the attack and the characteristics data; and
generate the set of candidate remediation plans based on different permutations of the remediation actions.
3. The system of claim 2 , wherein the remediation actions include one or more of:
applying a software patch to the set of machines,
disabling a feature provided by the set of machines,
removing one or more access privileges of one or more users from the set of machines,
updating one or more configuration settings of the set of machines, and
isolating a particular machine in the set of machines.
4. The system of claim 1 , wherein the remediation planning system is configured to display via the GUI estimated number of man hours needed to perform individual ones of the candidate remediation plans.
5. The system of claim 1 , wherein the remediation planning system is configured to display via the GUI estimated downtimes associated with performing individual ones of the candidate remediation plans.
6. The system of claim 1 , wherein the ranking is determined based at least in part on estimated number of man hours needed to individual ones of the candidate remediation plans or estimated downtimes associated with performing individual ones of the candidate remediation plans.
7. The system of claim 1 , wherein the remediation planning system is configured to:
continuously monitor the risk score of set of machines;
generate a detection event based on the monitoring of the risk score, wherein the detection event indicates the set of machines is vulnerable to the attack; and
display the detection event via the GUI, including a time of the detection event.
8. The system of claim 7 , wherein the detection event is generated as a user notification.
9. The system of claim 1 , wherein the GUI includes a control element to simulate the attack on the set of machines using an exploitation test system.
10. The system of claim 1 , wherein the exploitability risk model is a logistic regression model and trained to output a probability that the set of machines is exploitable by at least one of the set of attacks.
11. The system of claim 10 , wherein the exploitability risk model includes an attack likeliness model trained to output a probability of the set of machines being attacked based on machine demographic data, and an attack success model that outputs the risk score based at least in part on the probability of the set of machines being attacked.
12. The system of claim 11 , wherein the machine demographic data includes data collected by third-party machine monitoring services and includes data indicating prevalence of different machine characteristics and incidences of different attacks.
13. The system of claim 11 , wherein the exploitability risk model is retrained using results of penetration testing on the set of machines.
14. The system of claim 13 , wherein the exploitability risk model is retrained in response to changes in the characteristics data of the set of machines.
15. The system of claim 11 , wherein
the remediation planning system is implemented in platform-as-a-service provider network; and
the remediation planning system is configured to:
collect characteristics data of machine sets from a plurality of clients; and
recommend remediation plans for the machine sets using the exploitability risk model in response to detected attack vulnerabilities.
16. A method comprising:
performing, by one or more hardware processors with associated memory that implement a remediation planning system:
determining, using an exploitability risk model trained according to a machine learning technique, a risk score of a set of machines for an attack based on characteristics data of the set of machines;
determining a set of candidate remediation plans to reduce the risk score;
applying the candidate remediation plans to the characteristics data to generate updated instances of the characteristics data, wherein the application simulates performance of the candidate remediation plans on the set of machines without performing the candidate remediation plans on the set of machines;
determining, using the exploitability risk model and based on the updated instances of characteristics data, updated risk scores of the set of machines after the simulated performance of respective ones of the candidate remediation plans;
generating output via a graphical user interface (GUI) of the remediation planning system, wherein the GUI indicates, the attack, the risk score, the candidate remediation plans, the updated risk scores, and a ranking of the candidate remediation plans determined based at least in part on the updated risks scores; and
responsive to user selection of one of the candidate remediation plans received via the GUI, generating a workflow to perform the one candidate remediation plan on the set of machines.
17. The method of claim 16 , further comprising the remediation planning system:
selecting a plurality of remediation actions from a remediation action library based on the attack and the characteristics data; and
generating the set of candidate remediation plans based on different permutations of the remediation actions.
18. The method of claim 16 , further comprising the remediation planning system displaying via the GUI estimated number of man hours needed to perform individual ones of the candidate remediation plans.
19. The method of claim 16 , further comprising the remediation planning system displaying via the GUI estimated downtimes associated with performing individual ones of the candidate remediation plans.
20. The method of claim 16 , further comprising the remediation planning system launching an exploitation test system to simulate the attack on the set of machines in response to user input received via the GUI.