Accuracy in attack technique labeling in an extended detection system
Techniques for providing a confidence rating for an attack technique tag and a guidance for improving the confidence rating are described. An attack technique analytics engine receives telemetry data, a notification that an attack has been detected using an attack detection source, the notification including an attack technique tag indicating an attack technique used in the attack, and attack technique data including a likelihood of a particular attack technique, from among multiple attack techniques, occurring determined using a particular attack detection source from among multiple attack detection sources. Based on the telemetry, the attack technique tag, and the attack technique data, a confidence rating that the attack corresponds to the attack technique is determined. The attack detection source and attack technique are analyzed to determine a guidance for improving the confidence rating, and the rating and guidance are appended to the attack technique tag.
1 . A method performed at least in part by an attack technique analytics engine in an extended detection and response system, the method comprising:
receiving a stream of telemetry data generated by a telemetry source;
receiving, from an attack detection analytics engine, a notification that an attack technique has been detected using an attack detection source, the notification including an attack technique tag indicating an attack technique being used in the attack;
receiving, from an attack detection method scoring model, attack technique data including a likelihood of a particular attack technique, from among multiple attack techniques, occurring determined using a particular attack detection source from among multiple attack detection sources;
determining, based at least in part on (i) the stream of telemetry data, (ii) the attack technique tag, and (iii) the attack technique data, a confidence rating that the attack corresponds to the attack technique;
analyzing the attack detection source and the attack technique to determine a guidance for improving the confidence rating; and
appending the confidence rating and the guidance for improving the confidence rating to the attack technique tag.
2 . The method of claim 1 , wherein the attack technique is a first attack technique, the guidance is a first guidance, and the confidence rating is a first confidence rating and is below a predetermined threshold and further comprising:
determining, based at least in part on the attack detection source and the attack technique data, that the attack being detected is more likely to correspond to a second attack technique;
determining, based at least in part on (i) the stream of telemetry data, (ii) the second attack technique, and (iii) the attack technique data, a second confidence rating that the attack corresponds to the second attack technique;
analyzing the attack detection source and the second attack technique to determine a second guidance for improving the second confidence rating; and
generating a second attack technique tag that indicates the second attack technique, the second confidence rating, and the second guidance for improving the second confidence rating.
3 . The method of claim 2 , wherein the second attack technique tag includes the first attack technique with the first confidence rating and the second attack technique with the second confidence rating.
4 . The method of claim 3 , wherein the first confidence rating and the second confidence rating are within a threshold distance, and wherein the second guidance indicates an additional attack detection source for improving the first confidence rating or the second confidence rating.
5 . The method of claim 1 , wherein the stream of telemetry data is a first stream of telemetry data from a first telemetry source and the guidance for improving the confidence rating indicates that including a second stream of telemetry data from a second telemetry source would increase the confidence rating that the attack corresponds to the attack technique.
6 . The method of claim 1 , wherein the attack detection method scoring model is a machine learning model trained on datasets comprising an attack detection source used to accurately detect an attack technique and received from a technique database of know methods of attack detection.
7 . The method of claim 6 , wherein the technique database is an augmented MITRE ATT&CK Matrix.
8 . A system comprising:
one or more processors; and
one or more computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving, by an attack technique analytics engine, a stream of telemetry data generated by a telemetry source;
receiving, by the attack technique analytics engine and from an attack detection analytics engine, a notification that an attack has been detected using an attack detection source, the notification including an attack technique tag indicating an attack technique being used in the attack;
receiving, by the attack technique analytics engine and from an attack detection method scoring model, attack technique data including a likelihood of a particular attack technique, from among multiple attack techniques, occurring determined using a particular attack detection source from among multiple attack detection sources;
determining, by the attack technique analytics engine, based at least in part on (i) the stream of telemetry data, (ii) the attack technique tag, and (iii) the attack technique data, a confidence rating that the attack corresponds to the attack technique;
analyzing, by the attack technique analytics engine, the attack detection source and the attack technique to determine a guidance for improving the confidence rating; and
appending, by the attack technique analytics engine, the confidence rating and the guidance for improving the confidence rating to the attack technique tag.
9 . The system of claim 8 , wherein the attack technique is a first attack technique, the guidance is a first guidance, and the confidence rating is a first confidence rating and is below a predetermined threshold and the operations further comprising:
determining, based at least in part on the attack detection source and the attack technique data, that the attack being detected is more likely to correspond to a second attack technique;
determining, based at least in part on (i) the stream of telemetry data, (ii) the second attack technique, and (iii) the attack technique data, a second confidence rating that the attack corresponds to the second attack technique;
analyzing the attack detection source and the second attack technique to determine a second guidance for improving the second confidence rating; and
generating a second attack technique tag that indicates the second attack technique, the second confidence rating, and the second guidance for improving the second confidence rating.
10 . The system of claim 9 , wherein the second attack technique tag includes the first attack technique with the first confidence rating and the second attack technique with the second confidence rating.
11 . The system of claim 10 , wherein the first confidence rating and the second confidence rating are within a threshold distance, and wherein the second guidance indicates an additional attack detection source for improving the first confidence rating or the second confidence rating.
12 . The system of claim 8 , wherein the stream of telemetry data is a first stream of telemetry data from a first telemetry source and the guidance for improving the confidence rating indicates that including a second stream of telemetry data from a second telemetry source would increase the confidence rating that the attack corresponds to the attack technique.
13 . The system of claim 8 , wherein the attack detection method scoring model is a machine learning model trained on datasets comprising an attack detection source used to accurately detect an attack technique and received from a technique database of know methods of attack detection.
14 . The system of claim 13 , wherein the technique database is MITRE ATT&CK Matrix.
15 . One or more non-transitory computer-readable media storing instructions that, when executed, cause one or more processors to perform operations comprising:
receiving, by an attack technique analytics engine, a stream of telemetry data generated by a telemetry source;
receiving, by the attack technique analytics engine and from an attack detection analytics engine, a notification that an attack has been detected using an attack detection source, the notification including an attack technique tag indicating an attack technique being used in the attack;
receiving, by the attack technique analytics engine and from an attack detection method scoring model, attack technique data including a likelihood of a particular attack technique, from among multiple attack techniques, occurring determined using a particular attack detection source from among multiple attack detection sources;
determining, by the attack technique analytics engine, based at least in part on (i) the stream of telemetry data, (ii) the attack technique tag, and (iii) the attack technique data, a confidence rating that the attack corresponds to the attack technique;
analyzing, by the attack technique analytics engine, the attack detection source and the attack technique to determine a guidance for improving the confidence rating; and
appending, by the attack technique analytics engine, the confidence rating and the guidance for improving the confidence rating to the attack technique tag.
16 . The one or more non-transitory computer-readable media of claim 15 , wherein the attack technique is a first attack technique, the guidance is a first guidance, and the confidence rating is a first confidence rating and is below a predetermined threshold and the operations further comprising:
determining, based at least in part on the attack detection source and the attack technique data, that the attack being detected is more likely to correspond to a second attack technique;
determining, based at least in part on (i) the stream of telemetry data, (ii) the second attack technique, and (iii) the attack technique data, a second confidence rating that the attack corresponds to the second attack technique;
analyzing the attack detection source and the second attack technique to determine a second guidance for improving the second confidence rating; and
generating a second attack technique tag that indicates the second attack technique, the second confidence rating, and the second guidance for improving the second confidence rating.
17 . The one or more non-transitory computer-readable media of claim 16 , wherein the second attack technique tag includes the first attack technique with the first confidence rating and the second attack technique with the second confidence rating.
18 . The one or more non-transitory computer-readable media of claim 17 , wherein the first confidence rating and the second confidence rating are within a threshold distance, and wherein the second guidance indicates an additional attack detection source for improving the first confidence rating or the second confidence rating.
19 . The one or more non-transitory computer-readable media of claim 15 , wherein the stream of telemetry data is a first stream of telemetry data from a first telemetry source and the guidance for improving the confidence rating indicates that including a second stream of telemetry data from a second telemetry source would increase the confidence rating that the attack corresponds to the attack technique.
20 . The one or more non-transitory computer-readable media of claim 15 , wherein the attack detection method scoring model is a machine learning model trained on datasets comprising an attack detection source used to accurately detect an attack technique and received from a technique database of know methods of attack detection.