Spam Tide May Be Turning

Major announcements at the RSA Conference here last week—in addition to recent anti-spam technology advances—mark the beginning of the end of spam as we know it.

At the conference, Microsoft Corp. introduced its CSRI (Coordinated Spam Reduction Initiative), and Sendmail Inc. announced broad support of SMTP identification schemes.

Other anti-spam initiatives have moved ahead in recent weeks. The SPF (Sender Policy Framework), championed by Meng Weng Wong, gained traction on the news that it will be formally submitted to the Internet Engineering Task Force. Yahoo Inc.s Domain Keys, announced in December, has also bolstered the campaign for e-mail identity technology. Brightmail Inc.s Reputation Service and IronPort Systems Inc.s SMTPi initiative debuted late last month as well.

The premise of these new tools and initiatives is that once identity is effectively tied to e-mail messages, mail-handling systems will be able to forward legitimate e-mail and trash the forged junk now flooding the Internet.

eWEEK Labs therefore recommends that IT managers focus their energy on implementing new technology in their e-mail systems, instead of evaluating content-filtering anti-spam tools.

Because CSRI, SPF and other anti-spoofing technologies are still in the early stages of deployment, content-based anti-spam tools arent dead yet, of course. However, we believe IT managers should shift focus to participating in the pilot programs of e-mail identification systems and spend less time looking at the current crop of content-filtering tools.

E

-Mail ID Nitty-Gritty”>

E-Mail ID Nitty-Gritty

According to George Webb, group business manager at Microsoft, the first step in the CSRI framework, Caller ID, enables domain owners to assert their identity by adding records to their Domain Name System that allow recipients to verify the address of servers authorized to send e-mail.

“This is a technical proposal that, if adopted broadly across the e-mail infrastructure, would provide a great tool in fighting domain spoofing,” Webb said in a telephone interview last week.

The rest of CSRI boils down to a murkier set of mechanisms for senders to prove they are not spammers by using one of two methods. (The CSRI framework is at www.microsoft.com/mscorp/twc/privacy/spam.mspx.)

Large-volume senders will have to show they conform to rules and guidelines such as those contained in the federal CAN-SPAM legislation that became effective Jan. 1.

Senders can also buy their way into Bonded Sender, an IronPort program that is administered by TRUSTe. In an interview with eWEEK Labs, Tom Gillis, senior vice president of marketing at IronPort, explained IronPorts Bonded Sender program, in which users post a bond that is debited $10 every time a recipient complains of receiving spam. To date, no money has been paid out from Bonded Sender, Gillis said.

Individuals and small companies that cant afford to join programs like Bonded Sender could face Microsofts CSRI in the form of the Black Penny program. Black Penny is an anti-spam proposal that would require e-mail programs to process a difficult computational puzzle before e-mail could be sent. In effect, this would force the sender to burn CPU cycles that add cost to e-mail message generation.

These anti-spam proposals will go before the IETF and other standards bodies for ratification—but probably not until the specifications have become de facto industry norms. Now is the time for IT managers to get involved in the process.

It is likely that the technology choices made in the next year will set the direction of anti-spam efforts until (and if) SMTP can be completely ripped out and replaced with a reliable mail transport system.

Senior Analyst Cameron Sturdevant can be contacted at cameron_sturdevant@ziffdavis.com.