For a brief moment during the 12 months, it seemed like the boffins who invented smart software to defeat dumb problems had won the battle against the senders of electronic junk mail. The amount of spam actually received dropped dramatically between March and September, according to Internet researchers. Most people, armed with spam filters, got no more junk email than junk snail mail at their front door.
But in just three months, the spammers are reaching our inboxes again with their aggravating demands that we buy pirated software, usurious mortgages, and sex-enhancement pills and creams.
Either that - and this is the new, insidious wrinkle in Spam II - or a lot of your good email is going into the same reject folder as the spam. In the past 60 days, spammers have effectively defeated the email filters and trashed last year's efforts to make software that could identify and segregate the chaff from the wheat.
Now, the email babies are being thrown out with the spam dishwater. Many Internet users have already reverted to the painful task of looking at every incoming email to ensure that the "good" mail is not tossed away with the bad. The spam filter is on its last, teetering legs.
Here is a typical recent morning for one of the email accounts of a Bangkok Post reporter. There are 43 overnight messages in the spam-filtered email Inbox, of which 17 are spam. But there are 115 emails in the spam folder, of which three are legitimate email the reporter wants and needs.
Spammers have discovered that two systems are working, in order to get their junk through to email recipients - advertising embedded in images which spam filters can't "see", and hundreds of senseless, robot-written, boilerplate text emails, only a few of which actually are spam advertising.
The use of images in email is not new, but spammers have found it to be effective. They simply make a "picture" of their advertisement, and insert it into their email, just as millions of people send legitimate photos of families and events to each other every day.
Email security firm Postini said recently that image-embedded spam made up 45 percent of all spam this quarter, compared with about 2 percent a year ago.
No software filter currently exists which can accurately read a digital image and tell the difference, say, between an unpleasant sex aid advertisement and a family photo of a New Year's party.
The flood of text messages, however, is the newest wrinkle by the always aggressive, never apologetic direct-e-mail managers who are determined to feed you hundreds of advertisements no matter what you say about it.
These unpleasant people - experts say there are fewer than a dozen such men who control 85 per cent or more of the world's spam output - write simple computer programs to gather up thousands of words and sentences from web sites, word lists and news stories, and pack them together entirely randomly and senselessly.
The messages consist of real English (senseless to read, but then so are word lists). They contain no advertising at all, and like all spam these days are "from" entirely fake email addresses. There is nothing in these emails that absolutely identify them as spam. They are aimed specifically at crippling and eventually killing off the 2006 generation of anti-spam software which, just for a moment, seemed like it would rescue us from the electronic deluge of junk mail.
When a spam filter encounters such a message, whether on a major email website like Yahoo or Hotmail, or on a home computer, it eventually becomes confused over what is legitimate email and what is junk. Spam leaks to the In-box, important email is filtered to the spam folder.
The final nail in the current method of anti-spam defences is a hugely expanded system of "botnets" to send their spam ("bot" for "electronic robot").
Botnets are networks of computers which send spam - not only to Internet users, but to each other. As the mail bounces from machine to machine, there is no practical method of determining who is sending the spam, where he is based, or how to shut down his botnet.
Such networks are built mainly by hijacking the computers of regular Internet users. Ironically, these spam targets also are employed secretly to help to send out the spam. Computers are hijacked chiefly by using secretive adware, malware and intrusive software found at some web sites but more often within email messages. Experienced Internet users may know never to click on a web link in a suspect email, but tens of thousands of newcomers begin to use the world-wide web every day, and often without adequate software defences and Internet knowledge to safeguard their machines.
"The more high-speed connections and the more Windows PCs there are gives spammers that much more raw material," Postini executive vice president Dan Druker told one technical journal recently.
This makes the other, major method of blocking spam ineffective. Blacklisting Internet addresses no longer is reliable, because such action now is more likely to block an innocent web surfer than spam.
Two and a half years ago, Microsoft chairman Bill Gates bragged that his company and allies had "made significant advances against spam" and "spamming has become a more difficult and less rewarding business."
Whatever the fact of June 28, 2004, neither claim is either factual or true today.
Spamming is more automated and much easier today, and it is enormously rewarding. The only serious spammer shut down by toothless, worldwide legislation had a Hummer, a million-dollar mansion, $75,000 in cash - that is three million baht, just lying around the house and car - and a lifestyle that a successful yuppie could only dream of.
The other big "victory" was a civil lawsuit against a Colorado-based spam king, Scott Richter, who was known to have sent 38 billion junk emails containing 40,000 fraudulent statements, but certainly sent more. He had to promise never to do it again and pay a fine of $7 million, which he did swiftly on the day he was convicted, and of course never missed a meal.
Mr Gates has proposed a tiny fee to send email, a suggestion never to be adopted. Others have offered a one-year plan to change the entire, underlying system of sending, routing and receiving electronic mail, but this would be so disruptive - think changing traffic from the left side of the road to the right, but phasing it in over three months - that there only are two chances of its adoption in the foreseeable future: slim and fat.
Apart from these revolutionary stretches, the only automated anti-spam defence is likely to be a renewal of the cat-and-mouse game where spammers refine their current bombardment while software engineers sharpen and polish a toolbox of anti-spam software. Meanwhile, users for now are going to have to spend much more time in their spam folders, retrieving legitimate email from the dross.
In the medium future, the most effective anti-spam device will be education. Millions of Internet users have undefended computers. Spammers "enter" these machines and use them to send spam in the same way that thieves use a stolen cellphone to make long-distance calls until the number is cut off.
Internet users, like phone customers and home owners, will have to learn to use firewalls, hijack detectors and other methods of shutting down the worldwide spam network. |
