Systems and methods for identifying potentially malicious messages

Abstract

Computer-implemented systems and methods for identifying illegitimate messaging activity on a system using a network of sensors.

Description

TECHNICAL FIELD

This document relates generally to electronic communications processing and more particularly to analyzing electronic communications for spoofing and other situations.

BACKGROUND AND SUMMARY

A significant number of Internet users and companies are subject to spoofing attacks wherein an attacker masquerades as another person or company. An example includes a spoofing attack known as phishing wherein an attacker tries to illegally obtain confidential information (e.g., the user's password) by sending phony e-mails or instant messages and making the user believe that the source of the communication is a legitimate company. The technique is often used to try to secure user passwords and other sensitive information such as credit card numbers, bank account information, brokerage information and generally anything that could yield a financial gain in line with fraud operations.

In accordance with the teachings provided herein, systems and methods for operation upon data processing devices are provided in order to overcome one or more of the aforementioned disadvantages or other disadvantages concerning the detection of spoofing type situations. For example, a system and method can include examining whether an electronic communication includes elements associated with a first entity's website and elements associated with a second entity's website. The examination is then used in determining whether a spoofing situation exists with respect to the received electronic communication.

As another example, a computer-implemented method and system can be provided for detecting a spoofing situation with respect to one or more electronic communications, comprising. A determination is performed as to whether the electronic communication includes a textual or graphical reference to a first entity as well as a determination as to whether the textual or graphical reference to the first entity is associated with a link to a second entity. Spoofing is detected with respect to the received electronic communication based upon the determination of whether the textual reference is associated with the link to the second entity.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram depicting a computer-implemented system that includes a spoofed message detector to determine whether spoofing is evident with respect to one or more electronic communication messages.

FIG. 2 is a flowchart depicting operations that a message analysis system can utilize in determining the presence of spoofing.

FIG. 3 is a block diagram depicting a spoofed message detector configured to recognize a spoofed message.

FIG. 4 is a flowchart depicting an operational scenario for comparing two fingerprints of textual or graphical content.

FIG. 5 is a process flow diagram depicting application of a winnowing fingerprinting algorithm in order to detect spoofing.

FIG. 6 is a block diagram depicting a spoofed message detector configured to detect whether spoofing has occurred with respect to images.

FIG. 7 is a flowchart depicting an operational scenario for using fingerprint analysis in detecting spoofing.

FIG. 8 is a block diagram depicting a spoofed message detector configured to detect whether spoofing may have occurred with respect to communications that have direct links to a real website's images.

FIG. 9 is a flowchart depicting an operational scenario illustrating the analysis of direct links.

FIG. 10 is a block diagram depicting a spoofed message detector configured to be used with a reputation system.

FIGS. 11A and 11B are block diagrams illustrating actions that can be taken based upon the results of a spoofed message detector.

FIG. 12 is a block diagram depicting a server access architecture.

FIG. 13 is a block diagram depicting a message analysis system using an existing network of sensors.

FIG. 14 is a block diagram depicting the aggregation of threat data collected from existing sensors and external sources.

DETAILED DESCRIPTION

FIG. 1 depicts a computer-implemented system 30 that includes a spoofed message detector 32 to determine whether spoofing is evident with respect to one or more electronic messages (34, 36). As an example, the messages to be analyzed could be legitimate messages 34 from a company or could be spoofed messages 36 from an attacker feigning to be the company.

The legitimate messages 34 contain links to or elements from the company's website 38. The legitimate messages 34 can allow a recipient to access the company website 38 in order to perform a transaction or other activity through the company website 38. In contrast, spoofed messages 36 may contain links to or elements from the company's website 38 while also containing links to or elements from the attacker's website 40. This can result in the user being tricked into interacting with the attacker's website 40 instead of with the legitimate company's website 38.

The spoofed message detector 32 receives electronic communication (36, 34) over one or more networks 42. The spoofed message detector 32 analyzes the messages (36, 34) to determine whether spoofing may have occurred. If suspected spoofing has been detected with respect to an electronic message, then one or more actions 44 can take place with respect to the electronic communication. The actions 44 can be tailored based upon how likely the electronic communication is a spoofed message.

FIG. 2 represents operations that the message analysis system can utilize in determining the presence of spoofing. At step 100, a system can perform data collection to locate messages for analysis. For example, messages may be sent from devices that are located within one or more companies' networks. Such a device can include the IronMail message profiler device available from CipherTrust® (located in Alpharetta, Ga.).

From the data collected in step 100, step 102 determines which data is associated with which company. References to the company in the content, subject heading, and/or To/From/CC/BCC fields can be used to locate messages specific to a company. As an illustration, messages specific to Company A can be separated or otherwise indicated as being associated with Company A. Messages specific to Company B can be separated or otherwise indicated as being associated with Company B, and so forth. Other levels of granularity of separating the message can be performed, such as on an organization level, individual level, etc. In this manner, a user can direct analysis be performed at different levels of granularity.

Any messages that can be determined as legitimate at this stage can be removed from the corpus of messages that are to be analyzed at step 104. For example, messages can be determined as legitimate if their senders' addresses are from an advanced-authorized list of e-mail addresses, held by an ISP, subscriber or other e-mail service provider. At step 104, the remaining messages are analyzed to determine whether any of them are spoofed messages and if so, then one or more actions are performed at step 106 in order to address the spoofing situation.

A variety of different analysis techniques can be used to determine whether a spoofing situation has arisen at step 104, such as the approach depicted in FIG. 3. With reference to FIG. 3, the spoofed message detector 32 can be configured to recognize that a spoofed message 36 is a composite 200 of one or more elements 220 from a legitimate company website 38 as well as one or more elements 210 from a different entity's website (e.g., attacker's website 40). As an illustration of what website elements (210, 220) might be involved, the spoofed message detector 32 may detect that a message is a composite 200 because it includes content 222 from the legitimate company website 38 as well as content 212 from the attacker's website 40.

The spoofed message detector 32 can perform its composite analysis in many different ways. For example, the spoofed message detector 32 can utilize fingerprint analysis techniques 230 in order to determine whether the message is a composite 200 or not.

The spoofed message detector 32 can include or have access to a fingerprint analysis software routine or program 230 that will generate a fingerprint of the content 212 associated with a communication under analysis and generate a fingerprint of the actual content 220 used within the company website 38. A comparison of the two fingerprints generated by the fingerprint analysis program 230 is used to determine whether spoofing may have occurred. As operational scenario illustrating the use of fingerprinting analysis 230 is depicted in FIG. 4. It should also be understood that the fingerprinting analysis can be used to locate legitimate content. Such legitimate content can also be sorted for later analysis, such as, for example, trend analysis (e.g., how many times a legitimate usage is observed versus how many times a malicious usage is observed). Furthermore, it should be noted that instances of malicious usages can be stored for later use as evidence in a civil case or criminal case, or used in an administrative proceeding to shut malicious sites down.

With reference to FIG. 4, a communication to be analyzed is received by the spoofed message detector at process block 250. It should be understood that in various examples, the spoofed message detector can reside within an enterprise network, or any other generic location where messaging traffic may be observed. Moreover, when the spoofed message detector resides within an enterprise network, it should be noted that the detector can examine messaging traffic regardless of the originator of the message. For example, outgoing messages from the enterprise network may be examined to ensure that employees are not misusing the company mark or attempting to commit fraud with outsiders using company machines. Similarly, incoming messages maybe examined to protect employees from spoofing attacks by outsiders.

The spoofed message detector 32 identifies at process block 252 the different pieces of content referenced in the communication, such as what company-related content is being pointed to or hyperlinked in the communication. For example, a hyperlink in the communication might contain a textual description that indicates that it is a link to company content but instead provides a link to content on another website (e.g., an attacker's website)—this is an example of a communication faking an association with a company. The content is accessed and retrieved via the URL that is embodied in the hyperlink.

At process block 254, a fingerprint 256 is generated of the content that is actually pointed to or referenced in the communication that is under analysis. The fingerprint 256 is then made available to process block 262 which performs a comparison of fingerprint 256 with a fingerprint 260 that had been generated at process block 258. The comparison operation at process block 262 produces a matching result 264 indicative of how well the two compared fingerprints (256, 260) matched. A strong or complete match of the two fingerprints (256, 260) can provide evidence that spoofing has not occurred, while a partial match or a totally incomplete match can provide evidence that spoofing may be present.

As described above, it should be understood that various actions can be taken responsive to detecting suspected spoofing. For example, among others, suspected spoofing attacks can be added to a brand-abuse database, whereby messaging data can be combined with existing brand protection techniques.

It should be understood that similar to the other processing flows described herein, the steps and the order of the steps in this flowchart may be altered, modified and/or augmented and still achieve the desired outcome. For example, the generation of specific company's content fingerprint at process block 258 may be done in real-time or in near-real-time, such as when it has been discovered that the communication under analysis is referencing the specific company. The company content fingerprint could also be generated before the communication has been received for analysis. Furthermore, the comparison can use one or more techniques to determine whether a link or web page matches a legitimate link or web page.

As another example of the variety of processing flows that can be performed, the analysis does not have to include fingerprinting, but different comparison techniques can be utilized, such as a character-by-character comparison of the content involved in the analysis. Moreover, in various environments, different weightings can be applied to the different comparison techniques. If fingerprinting is utilized, then it should also be understood that different types of fingerprinting algorithms can be employed, such as the winnowing fingerprint algorithm discussed in the following reference: S. Schleimer et al. “Winnowing: Local Algorithms for Document Fingerprinting” (SIGMOD 2003, Jun. 9-12, 2003, San Diego, Calif.). An example of an application of the winnowing fingerprinting algorithm is shown in FIG. 5.

With reference to FIG. 5, a fingerprint of a “real” (i.e., authentic) login page of a company website is shown at 300. A fingerprint of the actual content that was contained in a communication purporting to be from the company is shown at 302. While many of the prints between fingerprints 300 and 302 may match, there are a number of significant departures between the two fingerprints 300 and 302. Accordingly, comparison operation 304 will produce a matching result 306 that would indicate that there is evidence of spoofing.

FIG. 6 depicts a spoofed message detector 32 that has been configured to detect whether spoofing may have occurred with respect to images 310 that have been incorporated into or is referenced by a message. Attackers may have downloaded images (e.g., company logos or other source indicating images, etc.) from the company's website. Accordingly, a spoofing situation could involve a composite 200 of website elements (e.g., images) from a company's website 38 as well as from an attacker's website 40. The spoofed message detector 32 can include or have access to a fingerprint analysis software routine or program 350 that will generate a fingerprint of an image 300 associated with a communication under analysis and generate a fingerprint of the actual image 310 used within the company website 38. A comparison of the two fingerprints generated by the fingerprint analysis program 350 is used to determine whether spoofing may have occurred.

As an illustration in detecting this type of spoofing, a company's images can be fingerprinted (e.g., by applying an md5 algorithm) and then these fingerprints can be compared against that of the communication in question or destination phishing website. Any matches not coming from the company's IPs can be deemed to be strong evidence of phishing. This could force phishers to modify their images which would result in more work for the phishers as well as increase the likelihood that people will not be fooled.

An operational scenario illustrating the use of fingerprint analysis 350 is depicted in FIG. 7. With reference to FIG. 7, an image is received at 400 that is associated with a communication to be analyzed. At step 402, the image from the company website is obtained. This image could have been obtained before or after the communication to be analyzed has been received.

At step 404, a fingerprint 406 of the image to be analyzed is generated. Correspondingly, at step 408, a fingerprint 410 of the company's image is generated. It is noted that the fingerprint 410 of the company's image could be generated before or after the communication to be analyzed is received.

The fingerprints 406 and 410 are then made available to process block 412 which performs a comparison of the fingerprints 406 and 410. The comparison operation at process block 412 produces a matching result 414 indicative of how well the two fingerprints (406, 410) matched. A strong or complete match of the two fingerprints (406, 410) can provide evidence that spoofing has not occurred, while a partial match or a totally incomplete match can provide evidence that spoofing may be present.

FIG. 8 depicts a spoofed message detector 32 that has been configured to detect whether spoofing may have occurred with respect to communications that have direct links 450 to the real website's images 310. An inventory of all the URLs belonging to a company can be performed periodically to reflect changes to a company's URLs. This inventory could be cross-referenced with a list of URLs permitted for real company communications. The inventory could also be cross-referenced with a fraud database in case any of the URLs that appear are not listed as officially belonging to the company. Any message that uses a mixture of real company URLs and fake URLs could be detected. Not only could this detect phishing but also trademark and other violations. If a phisher stops using valid company URLs, then message filters will be able to identify illegitimate mail, which would push phishers out into the open.

An operational scenario illustrating the analysis of direct links 450 is depicted in FIG. 9. With reference to FIG. 9, a communication is received at 500 that is to be analyzed in order to determine whether a spoofing situation (e.g., phishing) is present. At step 502, a list is generated of which company URLs are present in the communication. Either before or after the communication to be analyzed was received, process block 504 receives which URLs are allowed to be used for a company communication. Process block 506 does a comparison between the corpus of URLs obtained in process block 502 with the corpus of URLs obtained in process block 504. The comparison result 508 is indicative of whether spoofing has occurred.

FIG. 10 depicts a spoofed message detector 32 that has been configured to be used with a reputation system 550. A reputation system 550 keeps track of whether a communication sender engages in good behavior (such as sending legitimate messages 34), bad behavior (such as sending spam, malicious code, or spoofed messages 36). By tracking sender behavior over time, a database of sender reputation can grow and be refined.

Many different types of reputations system can be used with the spoofed message detector 32. An example includes the reputation systems and methods disclosed in the commonly assigned U.S. patent application entitled “Systems and Methods for Classification of Messaging Entities” (Ser. No. 11/142,943; filed Jun. 2, 2005). As another example, the spoofed message detector 32 can be used with a system, such as the TrustedSource software system provided by the assignee of this application. The TrustedSource software system receives and analyzes billions of messages per month from CipherTrust's network of more than 4000 IronMail Gateway appliances deployed globally. TrustedSource assigns a reputation score and further classifies senders as good, bad or suspicious based on an in-depth analysis by processing more than a dozen behavior attributes to profile each sender. As an illustration, TrustedSource combines traffic data, whitelists, blacklists and network characteristics with CiperTrust's global customer base.

The results of whether a message is a spoofed message can be provided to such reputation systems as part of its determination of what reputation should be ascribed to a particular sender. As an illustration, the determination by the spoofed message detector 32 (through one or more of the techniques disclosed herein) that a sender is sending spoofed messages can be used by a reputation system 550 to adversely affect the reputation of the sender.

As other examples of how the results of a spoofed message detector 32 can be used, FIG. 11A illustrates that an action 44 that can be taken based upon the results of the spoofed message detector 32 is to shutdown the attacker's website 40 as indicated at 600. The shutdown can be accomplished in a variety of ways, such to inform the Internet Service Provider (ISP) that the attacker's website 40 is associated with improper behavior (i.e., spoofing activities). Other ways could include a more automated approach to shutting down the attacker's website.

FIG. 11B illustrates that an action 44 could include modifications/alerts 650 being sent to the company 660 associated with the website 38. The company 660 is thereby aware of the illegitimate use of their identity and can decide what additional actions need to be taken. Additional actions could include pursuing legal action against the attacker, notifying persons (e.g., customers) to be aware of this phishing activity, etc.

While examples have been used to disclose the invention, including the best mode, and also to enable any person skilled in the art to make and use the invention, the patentable scope of the invention is defined by claims, and may include other examples that occur to those skilled in the art. For example, in addition to or in place of the other spoof message detection approaches discussed herein, a spoof message detector can be configured to determine whether a target/href mismatch has occurred in a communication under analysis. For example, a communication may indicate as its target http://www.ebay.com when it is really linking to http://215.32.44.3-ebay.com. Such a mismatch indicates that spoofing has occurred. This could be used in place of or to supplement the spoofing determinations performed by the other approaches discussed herein.

The systems and methods disclosed herein may be implemented on various types of computer architectures, such as for example on different types of networked environments. As an illustration, FIG. 12 depicts a server access architecture within which the disclosed systems and methods may be used (e.g., as shown at 30 in FIG. 12). The architecture in this example includes a corporation's local network 790 and a variety of computer systems residing within the local network 790. These systems can include application servers 720 such as Web servers and e-mail servers, user workstations running local clients 730 such as e-mail readers and Web browsers, and data storage devices 710 such as databases and network connected disks. These systems communicate with each other via a local communication network such as Ethernet 750. Firewall system 740 resides between the local communication network and Internet 760. Connected to the Internet 760 are a host of external servers 770 and external clients 780.

Local clients 730 can access application servers 720 and shared data storage 710 via the local communication network. External clients 780 can access external application servers 770 via the Internet 760. In instances where a local server 720 or a local client 730 requires access to an external server 770 or where an external client 780 or an external server 770 requires access to a local server 720, electronic communications in the appropriate protocol for a given application server flow through “always open” ports of firewall system 740.

A system 30 as disclosed herein may be located in a hardware device or on one or more servers connected to the local communication network such as on the Internet 760 and/or Ethernet 780 and logically interposed between the firewall system 740 and the local servers 720 and clients 730. Application-related electronic communications attempting to enter or leave the local communications network through the firewall system 740 are routed to the system 30.

System 30 could be used to handle many different types of e-mail and its variety of protocols that are used for e-mail transmission, delivery and processing including SMTP and POP3. These protocols refer, respectively, to standards for communicating e-mail messages between servers and for server-client communication related to e-mail messages. These protocols are defined respectively in particular RFC's (Request for Comments) promulgated by the IETF (Internet Engineering Task Force). The SMTP protocol is defined in RFC 821, and the POP3 is defined in RFC 1939.

Since the inception of these standards, various needs have evolved in the field of e-mail leading to the development of further standards including enhancements or additional protocols. For instance, various enhancements have evolved to the SMTP standards leading to the evolution of extended SMTP. Examples of extensions may be seen in (1) RFC 1869 that defines a framework for extending the SMTP service by defining a means whereby a server SMTP can inform a client SMTP as to the service extensions it supports and in (2) RFC 1891 that defines an extension to the SMTP service, which allows an SMTP client to specify (a) that delivery status notifications (DSNs) should be generated under certain conditions, (b) whether such notifications should return the contents of the message, and (c) additional information, to be returned with a DSN, that allows the sender to identify both the recipient(s) for which the DSN was issued, and the transaction in which the original message was sent. In addition, the IMAP protocol has evolved as an alternative to POP3 that supports more advanced interactions between e-mail servers and clients. This protocol is described in RFC 2060.

Other communication mechanisms are also widely used over networks. These communication mechanisms include, but are not limited to, Voice Over IP (VoIP) and Instant Messaging. VoIP is used in IP telephony to provide a set of facilities for managing the delivery of voice information using the Internet Protocol (IP). Instant Messaging is a type of communication involving a client which hooks up to an instant messaging service that delivers communications (e.g., conversations) that can take place in realtime.

FIG. 13 illustrates that some systems 30 of this disclosure operate using an existing network of sensors 800. In this example the sensors 800 are IronMail servers, publicly available from CipherTrust®, of Alpharetta, Ga. These sensors review mail traveling through associated network elements, such as mail transfer agents, for example. It should be understood that a user 805 creates a message and passes the message to an electronic mail server 810. A network 815a passes the message to a mail transfer agent which is associated with sensor 800. The sensor(s) 800 collects statistics related to messages reviewed and stores them in a database 820. The mail transfer agent forwards the mail to a recipient system 825 associated with a recipient of the message via a network 815b. It should be understood that the networks discussed herein can be the same network, or different subparts to the same network, although it should be understood that this disclosure is not limited to such an environment.

System 30 can examine the data stored by the sensor(s) 800 as described above. The system 30 can also make the data available to a client 835 (e.g., a web browser, and e-mail client, an SMS message, etc.) via a network 815c. In various examples, the client 835 can receive and/or retrieve information about potential spoofing activity. In the web-based example, a user could enter an IP address or domain name to observe the traffic associated with a system. In other examples, the detection system can send a message to a user or domain administrator, for example, via an ISP. Information can also be gathered from off-network areas, purchased from other companies and used for comparison and alert purposes within the system.

It should be further noted that the sensors 800 can gather information that would be useful to a company to determine whether anyone inside their company is transmitting illegitimate messaging traffic. Similarly, traffic patterns collected by the sensors 800 can be used to determine if there is concerted activity on the part of many computers associated with a domain or IP addresses. Such situations are evidence that a computer or network is infected with a virus, worm or spy-ware causing the computer or network to operate as a zombie client, thereby showing large increases in messaging traffic originating from a domain or IP address. Correlation of large amounts of messaging traffic indicates zombie activity, and helps administrators. Moreover, it can alert a reputation system to discount the messages sent by a domain or IP address during the period the system is influence by a zombie, a worm, or a virus, except where the problem persists (e.g., where the problem is ignored). An example of such a system is RADAR™, publicly available from CipherTrust®, which includes a customizable interface enabling users to configure notifications. CipherTrust also makes this information available via the web at: www.trustedsource.org. RADAR also includes a customizable interface to view messages and instances (indicated by URLs embedded in spoofed messages as well as URLs obtained from sources outside of the network of sensors) that indicate brand abuse—name, domain, website. Furthermore, the customizable interface can be configured in some examples to sort by one or more parameters such as, for example: sender, content, brand, time, location (corporate or geographic), among many others. Moreover, in some examples, data can be displayed in graphs, charts, and/or listed in tables, which enable the user to drill down to see different parts of the data (e.g., email header and/or entire message and content). Data from a geographical user interface (GUI) display can also be packaged for delivery (once or at regular intervals) in a file (which can be stored in any format including, for example: a text file, CSV file, a binary file, etc.). In various examples, views can be customized by user type or vertical type (e.g. an ISP view, or a Law Enforcement view, Banking view).

FIG. 14 illustrates an architecture 900 for aggregating data from a plurality of sensors 800a-c and external data received from other types of data collection systems such as data at rest. Data at rest can include, for example, among many others, the data stored on a domain name server or on a web server. It should be understood that each of the sensors 800a-C can include a local data store 820a-c, respectively, in which the sensor can store collected information. This data can be shared with system 30 via network(s) 815.

It should be understood that the stored data from the sensors 800a-c can be automatically sent to system 30, periodically, in times of low traffic or processor usage, or based upon some other triggering mechanism. Alternatively, the stored data from the sensors 800a-c can be automatically retrieved by the system 30, periodically, in times of low traffic or processor usage, or based upon some other triggering mechanism.

Additionally, system 30 can collect external data 905a-b, such as web data, domain name data, or other data at rest via the network(s) 815. The external data 905a-b can be collected by systems outside of the network of sensors. The external data 905a-b can be aggregated with the stored data received from the network of sensors 800a-c, as shown by aggregation block 910. The aggregated data can be sorted and/or analyzed as shown by block 920. The sorted and/or analyzed data can then be shared via the network(s) 815 using data server 930.

It should be understood that the data server can be used to provide the analyzed data to customers and other users via the world wide web, for example. Moreover, it should be noted that the sensors 800a-c can be configured to periodically retrieve the analyzed data from system 30, in order to operate on communication data using the latest threat and/or classification information to the sensors 800a-c.

It is further noted that the systems and methods may include data signals conveyed via networks (e.g., local area network, wide area network, internet, etc.), fiber optic medium, carrier waves, wireless networks, etc. for communication with one or more data processing devices. The data signals can carry any or all of the data disclosed herein that is provided to or from a device.

Additionally, the methods and systems described herein may be implemented on many different types of processing devices by program code comprising program instructions that are executable by the device processing subsystem. The software program instructions may include source code, object code, machine code, or any other stored data that is operable to cause a processing system to perform methods described herein. Other implementations may also be used, however, such as firmware or even appropriately designed hardware configured to carry out the methods and systems described herein.

The systems' and methods' data (e.g., associations, mappings, etc.) may be stored and implemented in one or more different types of computer-implemented ways, such as different types of storage devices and programming constructs (e.g., data stores, RAM, ROM, Flash memory, flat files, databases, programming data structures, programming variables, IF-THEN (or similar type) statement constructs, etc.). It is noted that data structures describe formats for use in organizing and storing data in databases, programs, memory, or other computer-readable media for use by a computer program.

The systems and methods may be provided on many different types of computer-readable media including computer storage mechanisms (e.g., CD-ROM, diskette, RAM, flash memory, computer's hard drive, etc.) that contain instructions for use in execution by a processor to perform the methods' operations and implement the systems described herein.

The computer components, software modules, functions, data stores and data structures described herein may be connected directly or indirectly to each other in order to allow the flow of data needed for their operations. It is also noted that a module or processor includes but is not limited to a unit of code that performs a software operation, and can be implemented for example as a subroutine unit of code, or as a software function unit of code, or as an object (as in an object-oriented paradigm), or as an applet, or in a computer script language, or as another type of computer code. The software components and/or functionality may be located on a single computer or distributed across multiple computers depending upon the situation at hand.

It should be understood that as used in the description herein and throughout the claims that follow, the meaning of “a,” “an,” and “the” includes plural reference unless that context clearly dictates otherwise. Also, as used in the description herein and throughout the claims that follow, the meaning of “in” includes “in” and “on” unless the context clearly dictates otherwise. Finally, as used in the description herein and throughout the claims that follow, the meanings of “and” and “or” include both the conjunctive and disjunctive and may be used interchangeably unless the context expressly dictates otherwise; the phrase “exclusive or” may be used to indicate situation where only the disjunctive meaning may apply.

Claims

1. A computer-implemented method for detecting a spoofing situation with respect to one or more electronic communications, the method comprising: receiving an electronic communication through a network interface addressed to a recipient;storing the electronic communication in computer memory;determining, by one or more processors and prior to the communication being provided to the recipient, that the electronic communication includes a link associated with a description of a first entity and that the link links to first content represented as particular content of the first entity, wherein the first content includes a first set of elements;identifying a legitimate version of the particular content including a second set of elements, wherein identifying the legitimate version includes identifying a first fingerprint of one or more elements from the legitimate version;generating a second fingerprint, wherein the second fingerprint is a fingerprint of one or more elements from the first content;determining a degree of match between the first and second sets of elements based at least in part on a comparison of the second fingerprint with the first fingerprint, wherein determining the degree of match includes determining whether one or more elements of the first set of elements originate from a second entity different from the first entity;detecting, by the one or more processors, prior to the communication being provided to the recipient that a spoofing situation exists with respect to the received electronic communication based upon the determined degree of match; andin response to detecting that a spoofing situation exists, blocking the communication from being provided to the recipient.

2. The method of claim 1, wherein the spoofing situation is a phishing situation wherein the link to the linked entity is a hyperlink to a website operated by the linked entity.

3. The method of claim 2, wherein the linked entity is an attacker whose website, to which the hyperlink links, is configured for feigning association with the first entity and for acquiring confidential information from a user for illegitimate gain.

4. The method of claim 1, wherein the first and second sets of elements include graphical elements.

5. The method of claim 1, wherein generating the fingerprints includes use of a winnowing fingerprint approach.

6. The method of claim 1, wherein the first and second sets of elements include image elements.

7. The method of claim 1, wherein the degree of match includes a match selected from the group consisting of a complete match, strong match, partial match, totally incomplete match, and combinations thereof.

8. The method of claim 1, further comprising: storing a number of instances of legitimate and illegitimate usages based upon whether the fingerprint of the first content and the fingerprint of the legitimate version of the particular content match; anddisplaying statistics comparing the number of instances of legitimate usage versus the number of instances of illegitimate usages.

9. The method of claim 1, wherein results of said detecting step are provided to a reputation system; wherein the reputation system uses the provided results as part of its determination of what reputation should be ascribed to a sender of the electronic communication.

10. The method of claim 1, wherein results of said detecting step are provided to a fraud database for correlation and aggregation.

11. The method of claim 1, wherein an action is performed in response to results of said detecting step; and wherein the action includes shutting down a website associated with the second entity.

12. The method of claim 1, wherein a notification is provided to the first entity of results of said detecting step.

13. The method of claim 1, further comprising: determining whether a mismatch has occurred with an href attribute in the received electronic communication; anddetecting whether a spoofing situation exists with respect to the received electronic communication based upon the determination with respect to the href attribute mismatch.

14. The method of claim 1, wherein the electronic communication is a communication selected from the group consisting of an e-mail message, and instant message, an SMS communication, a VOIP communication, a WAP communication, and combinations thereof.

15. The method of claim 1, further comprising: responsive to detecting a spoofing situation exists, performing at least one of the steps comprising changing the reputation of the sender of the communication.

16. The method of claim 1, wherein the step of detecting further comprises: determining a reputation associated with a URL included in the communication;determining whether the age of the domain used in the URL is greater than a threshold;determining whether the owner of the domain/IP hosting a URL included in the communication matches the owner of an IP address associated with the communication; anddetermining whether an owner of a phone number associated with the communication matches a database of known spoofing phone numbers.

17. The method of claim 1, wherein the first and second sets of elements include at least one of anchor text or an image.

18. A method of detecting illegitimate traffic originating from a domain, the method comprising: deploying a plurality of sensor devices at a plurality of associated nodes on the Internet;gathering messaging information from the plurality of sensor devices, wherein the messaging information describes messages originating from a set of domains including the domain;correlating a portion of the gathered messaging information for the domain;determining from the correlation whether a probable security condition exists with regard to the domain, wherein the determining the probable security condition comprises: comparing legitimate content of the domain with content contained in the gathered messaging information to identify that a volume of messages described in the received messaging information includes content that does not match the legitimate content, wherein the comparing includes: identifying a first fingerprint of one or more elements from the legitimate content;generating a second fingerprint, wherein the second fingerprint is a fingerprint of one or more elements from the content contained in the gathered messaging information; andcomparing the second fingerprint with the first fingerprint;signaling the probable security condition based at least in part on identifying that the volume of messages includes content that does not match the legitimate content; andalerting an owner or an internet service provider associated with the domain of the probable security condition with regard to the domain.

19. The method of claim 18, wherein the determining step comprises: comparing a list of company URLs contained in the gathered messaging information with an inventory of permitted URLs based upon one or more IP addresses associated with a particular entity associated with the domain; andif the list of company URLs contained in the received messaging information do not match the inventory of permitted URLs, signaling the probable security condition.

20. The method of claim 18, wherein the determining step comprises: comparing message traffic levels of multiple machines associated with the same domain, the messaging traffic levels being based on a messaging traffic level of messages from the domain; andif the message traffic levels of multiple machines associated with the samc domain display similar peak or similarly sporadic traffic levels during similar time periods, signaling the probable security condition.

21. The method of claim 18, wherein the sensor devices collect information about messaging traffic which travels across the associated nodes without regard to the origin or destination of the messaging traffic.

22. The method of claim 21, wherein the sensor devices collect information about all messaging traffic which travels across the associated nodes without regard to a protocol associated with the messaging traffic.

23. A method of detecting illegitimate traffic originating from a domain, the method comprising: deploying a plurality of sensor devices at a plurality of associated nodes on the Internet;gathering messaging information from the plurality of sensor devices, wherein the messaging information describes messages originating from a set of IP addresses including a particular IP address;correlating a portion of the gathered messaging information for the particular IP address;identifying a particular entity associated with the particular IP address;determining from the correlation whether a probable security condition exists with regard to the particular IP address, wherein the determining step comprises: comparing legitimate content of the particular entity with content contained in the gathered messaging information to identify that a volume of messages described in the received messaging information includes content that does not match the legitimate content, wherein the comparing includes: identifying a first fingerprint of one or more elements from the legitimate content;generating a second fingerprint, wherein the second fingerprint is a fingerprint of one or more elements from the content contained in the gathered messaging information; andcomparing the second fingerprint with the first fingerprint;signaling the probable security condition based at least in part on identifying that the volume of messages includes content that does not match the legitimate content; andalerting an owner associated with the particular IP address or an internet service provider associated with the IP address of the probable security condition with regard to the particular IP address.