Cyber security system and method using intelligent agents
OwnerRaymond Canfieldinventor
Life of this patent
Estimating the term…
Abstract
A cyber security method using intelligent agents (IAs) includes: watching, by the intelligent agent (IA), over a network, a software program running on a system; receiving, by the IA, results generated by the software; presenting, by the IA, the results; categorizing the results, by the IA, for efficient storage and efficient future retrieval; saving, by the IA, the categorized results; using the categorized results, by the IA, inferring new knowledge; categorizing the new knowledge, by the IA, for efficient storage and efficient future retrieval; saving, by the IA, the categorized new knowledge; and using one or more of the saved categorized results and the saved categorized new knowledge, by the IA, configuring the software.
Claims17
A cyber security method using intelligent agents (IAs), comprising:
- (a) after initial setup, directing, by an IA, a knowledge base program to scan data of a target;
- (b) collecting, by the IA, configuration information required to connect to the target to exploit a vulnerability;
- (c) using the configuration information, by the IA, attempting a connection to the target to exploit the vulnerability;
- (d) accumulating pertinent knowledge, by the IA, regarding one or more of a connection method and target information;
- (e) connecting, by the IA, with the target and exploiting the vulnerability with the target using the pertinent knowledge, wherein said exploiting the vulnerability with the target comprises binary level adjustments in the connection to the target to expose unknown vulnerabilities in the target, wherein the IA attempts the connection to a human user interface target by a world wide web (WWW) form request, wherein the IA completes the appropriate form request, and wherein the IA repeats the form request until the IA has no more requests available to make with the target;
- (f) receiving, by the IA, a response to a command from the target;
- (g) repeating steps (e) through (f) until the connection succeeds;
- (h) processing the response, by the IA, thereby generating a result;
- (i) transmitting, by the IA, the result to the knowledge base program;
- (j) using the knowledge base program, by the IA, processing the result;
- (k) receiving the processed result, by the IA, from the knowledge base program;
- (l) transmitting, by the IA, the processed result to a storage.
A cyber security method using intelligent agents (IAs) on a network, comprising:
- (a) after initial setup, directing, by an IA, a knowledge base program to continuously scan data of a target on a network;
- (b) using the configuration information, by the IA, attempting a connection to the target to exploit a vulnerability;
- (c) accumulating pertinent knowledge, by the IA, regarding one or more of a connection method and target information;
- (d) connecting, by the IA, with the target, to exploit the vulnerability with the target using the pertinent knowledge, wherein said exploiting the vulnerability with the target comprises binary level adjustments in the connection to the target to expose unknown vulnerabilities in the target, wherein the IA attempts the connection to a human user interface target by a world wide web (WWW) form request, wherein the IA completes the appropriate form request, and wherein the IA repeats the form request until the IA has no more requests available to make with the target;
- (e) collecting, by the IA, configuration information required to exploit the vulnerability with the target to exploit the vulnerability;
- (f) attempting, by the IA, to exploit the vulnerability of the target by repeatedly using the pertinent knowledge until the vulnerability is exploited;
- (g) receiving, by the IA, a response to a command from the target;
- (h) processing the response, by the IA, thereby generating a result;
- and
- (i) using a knowledge base program, by the IA, processing the results, and transmitting the processed result to a storage.
Dependencies are inferred from the claim text.