Hi,

This is for the one or two of you that might be interested:

SECURITY PIPELINE NEWSLETTER
Thursday, November 4, 2003

Elsewhere on the Internet, I gave a rather faint and
unenthusiastic endorsement to the federal CAN-SPAM legislation,
recently passed by both houses of U.S. Congress and expected to
be signed into law by President Bush.

I was never really wild about it. I thought it was a bad law, but
better than nothing. I was so un-enthusiastic about my support
that I actually headlined the piece, "CAN-SPAM Won't." The
article mainly wasn't about what's right with the law -- it was
about some of the things that are wrong with it.

Still, readers wrote back to let me know that I was talking
through my shorts even giving CAN-SPAM the grudging endorsement
that I gave it. CAN-SPAM isn't better than nothing -- it's at
best nothing, and it might even do more harm than good.

Let me explain. The CAN-SPAM act has several requirements
designed -- or so the bill says -- to limit spam. Senders are
required to provide real, working unsubscribe instructions in the
newsletter. Senders are prohibited from using misleading headers,
including subject lines, to hide the source and contents of their
messages. And the Federal Trade Commission is required to study
the feasibility of setting up a "do not spam" list, similar to
the existing do-not-call list for telemarketers.

In my earlier article, I said that the requirement to provide
unsubscribe instructions was useless because recipients would be
overwhelmed by the need to unsubscribe one at a time from a flood
of incoming spam. However, I was hopeful that the prohibition
against misleading headers might at least make it practical for
e-mail administrators to set up real, working blacklists, instead
of the flawed blacklists we have today. And I was also hopeful
that a do-not-spam list might work.

Well, the readers let me know that I was making my assumptions
based on a flawed understanding of how e-mail works.

Reader Xavier Haurie explained:

So how does this piece of legislation work when you TECHNICALLY
can't verify the identity of the sender? What's really missing is
an official body and funds to define and test a new standard for
identifying email senders....

I don't doubt that you know that authentication is not a problem
in the case of telemarketing because (1) telephone works by
closing a circuit b/w 2 telephones with known numbers and (2) the
telephone companies knows who owns and operates those phones.
Whereas an email is like a glorified IP data packet, or a letter
in the mail. It'll get delivered so long as postage is paid and
the destination address is correct.

In other words, forbidding senders from falsifying headers only
works if you have some way of figuring out who sent the mail in
the first place. Telemarketing can be regulated because
authentication is built into the phone system -- you can't run a
telemarketing operation without leaving some record of who you
called and when you placed the call. Whereas it's trivially easy
to send anonymous e-mail, and very difficult to prove who sent an
e-mail.

Even if a company -- say, Spacely Sprockets -- sends a spam
message with its name and phone number attached to it, when the
cops come to call all they have to do is deny sending it. "My
enemies sent it to cause me to run afoul of the CAN-SPAM act. It
was those dastardly villains at my competitors, Cogswell Cogs,"
they'll say, and the cops won't be able to prove differently.

Likewise, a do-not-spam list would simply be a massive list of
real, working e-mail addresses which the spammers would love to
get their hands on.

Also, the law does nothing to address the problem of overseas
spammers. It does not allow for private action -- that is, if you
receive a spam, you can't sue the spammer, you have to send a
letter to your state attorney general and wait for THEM to sue.

That's why CAN-SPAM is, at best, useless.

At worst, CAN-SPAM is harmful. It would supersede more effective
state legislation, such as a proposed law in California.
Moreover, it wouldn't BAN spam, it would, on the contrary,
LEGALIZE it -- marketers would know that, if they follow the
guidelines of CAN-SPAM, they can send spam messages with
impunity.

As reader Haurie says, the real fix for spam is to replace or
extend e-mail protocols so that senders can be authenticated.
Sure, it's important to preserve anonymity on the Internet, to
allow the Internet to be used for political dissent,
psychological support groups, and for confidential medical
research. But most of the time, we need to be able to identify
who is trying to contact us. Those two goals are compatible --
and anonymity is, now, killing the Internet.

P.S. Analysts Gartner today warned corporations that CAN-SPAM
won't help slow down spam, but it will result in greater scrutiny
of e-mail messages.

URLs referenced in this item:

CAN-SPAM Won't
http://wagblog.internetweek.com/archives/000804.html

Gartner To Corporations: Don't Rely On CAN-SPAM
http://www.securitypipeline.com/news/showArticle.jhtml?articleId=16506458

Also:

Do Not Bother With A Do Not Spam List
http://www.securitypipeline.com/story/showArticle.jhtml?articleID=15500142

-- Mitch Wagner, Co-Editor, Security Pipeline
http://www.securitypipeline.com/
[EMAIL PROTECTED]

--
Assured Computing
When you need to be sure.
[EMAIL PROTECTED]
www.assuredcomp.com
Voice - 541-868-0331
FAX - 541-463-1627
Eugene, Oregon


_______________________________________________
EuG-LUG mailing list
[EMAIL PROTECTED]
http://mailman.efn.org/cgi-bin/listinfo/eug-lug

Reply via email to