Wednesday, January 7, 2009

Filtering junk mail costs a packet

Globally, junk email or spam has grown, at times exponentially, to reach over 100 billion messages a day in April 2008.

This may surprise some users because the amount of received spam has decreased, mostly thanks to better filtering methods. Still, the cost of processing and removing this data is eventually borne by users as internet providers pass it on. We have an array of servers both in Cape Town and Houston dedicated to filtering junk mail and scanning for virii.

Email addresses are collected from internet chat rooms, websites, newsgroups, and viruses which "harvest" address books in popular email software.

Databases of these addresses are then sold to other spammers - some of whom are multimillionaires thanks to their efforts. Loads of spam are sent to invalid email addresses, causing congestion and wasted capacity - all costs to internet providers and their customers.

Our array of Cape Connect mail servers reject messages for non existant email accounts on an ongoing basis.

What's arguably even more worrying is how the spam is sent. Using a computer virus, spammers can infect computers and turn them into "zombie PCs", churning out thousands of emails every day without the user's permission or even their knowledge.

In late 2006, more than 80% of spam was distributed in this way. Spammers love this technique as it takes minimal resources, and is harder to prosecute as the origin of the spam is difficult to trace.

Group e-mails among friends are a prime example of how carelessness or ignorance on the part of a few email users can contribute to the spam problem. I'm referring to messages sent to multiple recipients, all listed in the cc: (carbon copy) header - with all their addresses visible to each and every other recipient.

It's important to realise the implications of using the carbon copy field - especially if you're not certain of a recipient computer's security software, or how up to date it is. I regularly cite this example of a group email I received some time ago:

Someone selling a 4x4 privately, decided to send by email a sizeable portion of his email address book.

Let's call this person A. A's mail listed about 120 email addresses. I didn't receive the original mail, and I don't know A. However, I received the mail because a mutual friend of A's and mine, who was on the original list of addresses, in turn forwarded it to everyone he knew, in the process adding more than two hundred addresses from his address book. Let's call this mutual friend B.

Presumably the truck was a good deal, because before long, other mutual friends of B and mine also forwarded the mail. My address appeared in their address books too - so I got the message several times again.

Then it started to get interesting. Some of the recipients of this forwarded message, who I'd never met, are "Switched On" readers - and they, perhaps unwittingly, forwarded the message to its address, through the same "select all" approach.

Then a local 4x4 accessories shop seized the opportunity and sent out an advert for their full range of goods and services.

Soon a competitor did the same, this time adding a huge attachment. Some recipients objected to this advert and complained by "replying to all" - further exacerbating the issue. Within a week of A's original mail, I'd received the advert for the truck 11 times, and seven other related emails, from a variety of sources. In the process, I also managed to collate more than 900 different email addresses - and that's where the real problem lies...

Our Main Site | Our Classified Adverts