Detection and prevention of external fraud
Techniques for detecting instances of external fraud by monitoring digital activities that are performed with accounts associated with an enterprise are disclosed. In one example, a threat detection platform determines the likelihood that an incoming email is indicative of external fraud based on the context and content of the incoming email. To understand the risk posed by an incoming email, the threat detection platform may seek to determine not only whether the sender normally communicates with the recipient, but also whether the topic is one normally discussed by the sender and recipient. In this way, the threat detection platform can establish whether the incoming email deviates from past emails exchanged between the sender and recipient.
1 . A system, comprising:
a processor configured to:
obtain an email that is addressed to an email account associated with an enterprise;
examine the email so as to identify a domain from which the email originated;
determine, based on the domain, whether the email is representative of an attempt to impersonate a vendor including by being configured to: access a federated vendor database comprising a plurality of digital profiles, wherein each digital profile corresponds to a different legitimate vendor and is learned through analysis of past emails addressed to email accounts associated with a plurality of different enterprises, and compare the domain to legitimate domains included in the digital profiles; and
prohibit the email from reaching an inbox of the email account responsive to a determination that the email is representative of an attempt to impersonate the vendor; and
a memory coupled to the processor and configured to provide the processor with instructions.
2 . The system of claim 1 , wherein said determining comprises:
applying, to the email, a model that is trained to identify instances of impersonation based on an analysis of (i) content of the email, (ii) context of the email, or (iii) age of the domain; and
establishing whether the email is representative of an attempt to impersonate the vendor based on an output produced by the model.
3 . The system of claim 1 , wherein said determining comprises comparing the domain to digital profiles maintained in a database, wherein each digital profile includes a legitimate domain of a different vendor learned through analysis of past emails.
4 . The system of claim 3 , wherein the past emails are addressed to email accounts associated with the enterprise and at least one other enterprise.
5 . The system of claim 1 , wherein the processor is further configured to:
access a database that includes information regarding registered domains so as to establish an age of the domain;
wherein said determining is further based on the age of the domain.
6 . The system of claim 1 , wherein the processor is further configured to:
access a database that includes information regarding registered domains so as to establish an age of the domain; and
wherein said determining comprises:
inferring an intent of the email based on an analysis of (i) content of the email, (ii) context of the email, and (iii) the age of the domain; and
establishing whether the email is representative of an attempt to impersonate the vendor based on the intent.
7 . The system of claim 1 , wherein the processor is further configured to:
indicate in a digital profile associated with the vendor that the vendor has been a target of an impersonation attempt;
produce a score that is indicative of a threat posed by communicating with the vendor; and
store the score in the digital profile associated with the vendor.
8 . The system of claim 7 , wherein said indicating comprises storing the email in the digital profile associated with the vendor.
9 . A method, comprising:
obtaining an email that is addressed to an email account associated with an enterprise;
examining the email so as to identify a domain from which the email originated;
determining, based on the domain, whether the email is representative of an attempt to impersonate a vendor including by: accessing a federated vendor database comprising a plurality of digital profiles, wherein each digital profile corresponds to a different legitimate vendor and is learned through analysis of past emails addressed to email accounts associated with a plurality of different enterprises, and comparing the domain to legitimate domains included in the digital profiles; and
prohibiting the email from reaching an inbox of the email account responsive to a determination that the email is representative of an attempt to impersonate the vendor.
10 . The method of claim 9 , wherein said determining comprises:
applying, to the email, a model that is trained to identify instances of impersonation based on an analysis of (i) content of the email, (ii) context of the email, or (iii) age of the domain; and
establishing whether the email is representative of an attempt to impersonate the vendor based on an output produced by the model.
11 . The method of claim 9 , wherein said determining comprises comparing the domain to digital profiles maintained in a database, wherein each digital profile includes a legitimate domain of a different vendor learned through analysis of past emails.
12 . The method of claim 11 , wherein the past emails are addressed to email accounts associated with the enterprise and at least one other enterprise.
13 . The method of claim 9 , further comprising accessing a database that includes information regarding registered domains so as to establish an age of the domain; wherein said determining is further based on the age of the domain.
14 . The method of claim 9 , further comprising:
accessing a database that includes information regarding registered domains so as to establish an age of the domain; and
wherein said determining comprises:
inferring an intent of the email based on an analysis of (i) content of the email, (ii) context of the email, and (iii) the age of the domain; and
establishing whether the email is representative of an attempt to impersonate the vendor based on the intent.
15 . The method of claim 9 , further comprising:
indicating in a digital profile associated with the vendor that the vendor has been a target of an impersonation attempt;
producing a score that is indicative of a threat posed by communicating with the vendor; and
storing the score in the digital profile associated with the vendor.
16 . The method of claim 15 , wherein said indicating comprises storing the email in the digital profile associated with the vendor.
17 . A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:
obtaining an email that is addressed to an email account associated with an enterprise;
examining the email so as to identify a domain from which the email originated;
determining, based on the domain, whether the email is representative of an attempt to impersonate a vendor including by: accessing a federated vendor database comprising a plurality of digital profiles, wherein each digital profile corresponds to a different legitimate vendor and is learned through analysis of past emails addressed to email accounts associated with a plurality of different enterprises, and comparing the domain to legitimate domains included in the digital profiles; and
prohibiting the email from reaching an inbox of the email account responsive to a determination that the email is representative of an attempt to impersonate the vendor.
18 . The computer program product of claim 17 , wherein said determining comprises:
applying, to the email, a model that is trained to identify instances of impersonation based on an analysis of (i) content of the email, (ii) context of the email, or (iii) age of the domain; and
establishing whether the email is representative of an attempt to impersonate the vendor based on an output produced by the model.
19 . The computer program product of claim 17 , wherein said determining comprises comparing the domain to digital profiles maintained in a database, wherein each digital profile includes a legitimate domain of a different vendor learned through analysis of past emails.
20 . The computer program product of claim 19 , wherein the past emails are addressed to email accounts associated with the enterprise and at least one other enterprise.