Phishing email campaign identification

A computer-implemented method for identifying phishing emails from a malicious email campaign, the method comprising, executed by one or more email detection computers:

• receiving, from a digital electronic computer network, a first email message that is directed to a first recipient account from a first sender account, the first email message being associated with a plurality of attributes;
• determining that the first email message is a phishing email;
• extracting, from the first email message, a subset of the plurality of attributes, the subset of the plurality of attributes comprising a plurality of fixed attributes and a plurality of transformable attributes, the plurality of fixed attributes of the first email message being two or more of a thread count;
• an attachment count;
• a TO recipient count;
• a CC recipient count;
• a number of URLs within a body of the first email message;
• a number of links within the body of the first email message;
• or a BCC recipient count;
• transforming each of the plurality of transformable attributes to a normalized representation of the transformable attribute;
• concatenating the plurality of fixed attributes and the normalized representations of the plurality of transformable attributes into a string representation;
• generating a hash representation of the subset of the plurality of attributes based on the string representation of the concatenated plurality of fixed attributes and the normalized representations of the plurality of transformable attributes;
• storing the hash representation of the subset of the plurality of attributes of the first email message in a database;
• receiving, from the digital electronic computer network, a second email message that is directed to a second recipient account from a second sender account;
• determining that the second email message is a phishing email based on the hash representation corresponding to the first email message and by:
• extracting, from the second email message, a second subset of the plurality of attributes, the second subset of the plurality of attributes comprising a second plurality of fixed attributes and a second plurality of transformable attributes;
• transforming each particular transformable attribute of the second plurality of transformable attributes to a second normalized representation of that particular transformable attribute;
• generating a second hash representation of the subset of the second plurality of attributes based on the second plurality of fixed attributes and the second normalized representations of the second plurality of transformable attributes;
• comparing the second hash representation of the second subset of the second plurality of attributes corresponding to the second email message with the hash representation corresponding to the first email message;
• in response to determining that the second email message is the phishing email based on the hash representation corresponding to the first email message, executing a particular responsive action from among a plurality of responsive actions comprising blocking forwarding the second email message, moving the second email message to a quarantine, transmitting an alert toward the first recipient account, dropping or deleting the second email message.

One or more non-transitory computer-readable storage media including instructions that, when executed by one or more processors, are configured to cause the one or more processors to:

• receive, from a digital electronic computer network, a first email message that is directed to a first recipient account from a first sender account, the first email message being associated with a plurality of attributes;
• determine that the first email message is a phishing email;
• extract, from the first email message, a subset of the plurality of attributes, the subset of the plurality of attributes comprising a plurality of fixed attributes and a plurality of transformable attributes, the plurality of fixed attributes of the first email message being two or more of a thread count;
• an attachment count;
• a TO recipient count;
• a CC recipient count;
• a number of URLs within a body of the first email message;
• a number of links within the body of the first email message;
• or a BCC recipient count;
• transform each particular transformable attribute of the plurality of transformable attributes to a normalized representation of that particular transformable attribute;
• concatenate the plurality of fixed attributes and the normalized representations of the plurality of transformable attributes into a string representation;
• generate a hash representation of the subset of the plurality of attributes based on the string representation of the concatenated plurality of fixed attributes and the normalized representations of the plurality of transformable attributes;
• store the hash representation of the subset of the plurality of attributes of the first email message in a database;
• receive, from the digital electronic computer network, a second email message that is directed to a second recipient account from a second sender account;
• determine that the second email message is a phishing email based on the hash representation corresponding to the first email message.

One or more non-transitory computer-readable including instructions that, when executed by one or more processors, are configured to cause the one or more processors to:

• receive, from a digital electronic computer network, a first email message that is directed to a first recipient account from a first sender account, the first email message being associated with a plurality of attributes;
• determine that the first email message is a phishing email;
• extract, from the first email message, a subset of the plurality of attributes, the subset of the plurality of attributes comprising a plurality of fixed attributes and a plurality of transformable attributes, the plurality of fixed attributes of the first email message being two or more of a thread count;
• an attachment count;
• a TO recipient count;
• a CC recipient count;
• a number of URLs within a body of the first email message;
• a number of links within the body of the first email message;
• or a BCC recipient count;
• transform each of the plurality of transformable attributes to a normalized representation of the transformable attribute;
• concatenate the plurality of fixed attributes and the normalized representations of the plurality of transformable attributes into a string representation;
• generate a hash representation of the subset of the plurality of attributes based on the string representation of the concatenated plurality of fixed attributes and the normalized representations of the plurality of transformable attributes;
• store the hash representation of the subset of the plurality of attributes of the first email message in a database;
• receive, from the digital electronic computer network, a second email message that is directed to a second recipient account from a second sender account;
• determine that the second email message is a phishing email based on the hash representation corresponding to the first email message and by:
• extracting, from the second email message, a second subset of the plurality of attributes, the second subset of the plurality of attributes comprising a second plurality of fixed attributes and a second plurality of transformable attributes;
• transforming each particular transformable attribute of the second plurality of transformable attributes to a second normalized representation of that particular transformable attribute;
• generating a second hash representation of the subset of the second plurality of attributes based on the second plurality of fixed attributes and the second normalized representations of the second plurality of transformable attributes;
• comparing the second hash representation of the second subset of the second plurality of attributes corresponding to the second email message with the hash representation corresponding to the first email message;
• in response to determining that the second email message is the phishing email based on the hash representation corresponding to the first email message, execute a particular responsive action from among a plurality of responsive actions comprising blocking forwarding the second email message, moving the second email message to a quarantine, transmitting an alert toward the first recipient account, dropping or deleting the second email message.