As spammers continue to come up with news ways to reach more e-mail users with their unwanted commercial messages, security
vendors keep working on new techniques for blocking spam. On Monday, CipherTrust and Cloudmark each announced new twists to
their spam-fighting measures.
CipherTrust, slated to become a part of Secure Computing on September 8 once the announced acquisition plans are approved, has added new fingerprinting techniques to its TrustedSource reputation system that is part of its IronMail messaging gateway appliance, says Paul Judge, CTO of CipherTrust.
The company has added fingerprinting techniques, which break a document down into components and assign unique alphanumeric
identifiers to each portion so they can be easily identified, giving the reputation system another way to spot spam. Currently
the system identifies IP addresses known to have sent spam and can block e-mail coming from them for a certain length of time;
with fingerprinting, the system can now also check inbound mail to see if it matches known spam messages that have been fingerprinted,
and block them accordingly.
“Once we’ve seen messages from known bad senders, when we see [the same message] coming from unknown senders we know they’re
spam,” says Judge.
The new feature, called TrustedSource Message Reputation, analyzes thousands of fingerprints for each inbound message, including
any embedded images and attachments, looking for a match. This method can be particularly helpful in combating image-based
spam, or messages in which the text is embedded in an image and therefore can’t be read by traditional filters, which has
jumped from 10% of all spam messages in May to 30% currently, says Judge.
