[IP] mo on Spammers Most Likely Users of Email Authentication
-----Original Message-----
From: "Suresh Ramasubramanian"<suresh@xxxxxxxxxx>
Sent: 13/07/05 7:04:41 AM
To: "dave@xxxxxxxxxx"<dave@xxxxxxxxxx>
Cc: "Randall"<rvh40@xxxxxxxxxxxxx>
Subject: Re: [IP] Spammers Most Likely Users of Email Authentication
Lots of large ISPs are in fact removing their SPF records - we
(outblaze.com) did so some months back, and were probably the first
large site to do this publicly. Now, some other large providers are
also doing this.
I wrote about why we did this in this circleid article -
http://www.circleid.com/article/1039_0_1_0_C/ (there is an interesting
thread of comments in that article, evaluating the other proposals as well)
DKIM looks promising - but again, like spf or sender id, it is not a
cure for spam. It is just a way to stop forged email.
Authentication schemes like these just say, for example "This email says
it came from joebloggs@xxxxxxxxxxx - and seeing that the authentication
checks out, it indeed did come from joebloggs@xxxxxxxxxxx [or at least
from example.com]".
Kind of like your walking into a movie theater to see a movie, the
ticket says "The War of the Worlds", the sign outside the theater says
"War of the Worlds", chances are that you are going to see "War of the
Worlds"
Reputation services on the other hand are like Roger Ebert's movie
reviews, or blog posts or whatever. "War of the Worlds? What a lousy
movie", or "Hey, great movie, you really got to go see it".
Authentication therefore has to go hand in hand with reputation
Where some authentication schemes become quite popular is "edge" cases
where they get used.
One currrent favorite edge case is that Hotmail says they'll tag mail
failing sender id checks with what keeps getting called a BYBOD - "Big
Yellow Box Of Doom".
And AOL says that email sending sites (ISPs, bulk mailers etc) can
automatically update their whitelist with AOL using SPF records to
designate their sending IPs, without having to follow the "manual"
method of opening a ticket with AOL's postmaster staff and telling them
about any new changes in your IP space so they can manually update their
whitelist.
Large domain speculators who register thousands of domains but dont want
to run mailservers for those till they are sold will set up a bogus MX
record like dev.null in DNS and publish a "v=spf1 -all" record to say
"this domain doesnt send any email at all" [this is by the way a great
way to turn off email service entirely for a domain that you know wont
need a mailserver]
etc etc
SPF and other auth schemes may be quite useful in ways that are not
really an intended use of the spec :)
regards
-suresh
David Farber wrote:
>
>
> Begin forwarded message:
>
> From: Randall <rvh40@xxxxxxxxxxxxx>
> Date: July 12, 2005 6:27:31 PM EDT
> To: Dave <dave@xxxxxxxxxx>
> Subject: Spammers Most Likely Users of Email Authentication
>
>
> http://tinyurl.com/cnltn
>
> Spammers Most Likely Users Of E-Mail Authentication
> By Gregg KeizerMon Jul 11, 3:40 PM ET
>
> On the eve of an industry summit to discuss how e-mail authentication
> can stem the flood of spam, one security firm says that spammers are
> already using the protocols -- to slip their junk mail past filters.
>
> According to Denver-based message security vendor MX Logic, spammers are
> continuing to adopt Sender ID and Sender Policy Framework (SPF), two of
> the prominent e-mail authentication schemes that are actually intended
> to stop spam.
>
> MX Logic tracked a sampling of 17.7 million messages that passed through
> its servers from June 19 through June 25, and found that of the 9
> percent from domains with published SPF records, 84 percent was spam. Of
> the even smaller number of messages from domains with published Sender
> ID records (just 0.14 percent), 83 percent were spam.
>
> "Spammers continue to leverage SPF and Sender ID with the intention of
> making their messages appear more legitimate," said Scott Chasin, MX
> Logic chief technology officer, in a statement.
>
> All e-mail authentication schemes work in similar fashion -- by relying
> on DNS records, either to obtain sending mail server addresses or public
> keys for decrypting a digital signature -- and are designed to reduce or
> prevent "spoofing," the spammer and phisher tactic of forging e-mail
> From: addresses.
>
> Microsoft recently reworked its free-of-charge, Web-based Hotmail
> service so that all messages not using Sender ID are identified as such.
> The Redmond, Wash.-based developer isn't, however, deleting non-Sender
> ID mail or trashing it by placing it in a junk mail filter. Yet.
>
> "As adoption of Sender ID and SPF records grows, and the lack of a
> domain with an SPF record becomes the exception to the norm, we may
> choose to investigate unauthenticated e-mail more closely before
> deciding whether to deliver it to the users' inbox," said Craig Spiezle,
> http://www.microsoft.com/presspass/features/2005/
> jun05/06-22SenderID.mspx a director in Microsoft's safety group, when
> the move was announced in late June.
>
> MX Logic's Chasin saw a connection between spammers using SPF and Sender
> ID and the Hotmail move. "[Spammers are trying to] avoid having their
> messages delivered with an onscreen notification that a Sender ID record
> was not found, a method Microsoft recently announced it will use on
> Hotmail," he wrote.
>
> That spammers have rushed to adopt sender authentication isn't the only
> concern for the protocols. On Monday, the Message Anti-Abuse Working
> Group, an industry group that includes AOL, Yahoo, Symantec, and
> EarthLink, released a report detailing a six-month evaluation of SPF and
> Sender ID. The report acknowledged problems with the protocols when
> forwarding or re-sending mail, and noted that no authentication scheme
> can guarantee a message really does come from who it says it came from.
>
> "At best, SPF and Sender ID are comparable to a license plate issued by
> a foreign country: they show that the vehicle is permitted to drive in
> that country, but make no indication as to whether that country's
> regulations are similar to yours, and we can only assume that the driver
> inside is permitted to use that vehicle," the report concluded.
>
> Also on Monday, Yahoo and Cisco submitted their combined authentication
> standard, DomainKeys Identified Mail (DKIM) to the Internet Engineering
> Task Force (IETF) for consideration as a standard. The DKIM protocol,
> which borrowed bits from both Yahoo's DomainKeys and Cisco's Internet
> Identified Mail, was first announced in early June.
> http://informationweek.com/story/showArticle.jhtml?articleID=163702984
> It's expected that DKIM will be discussed at the IETF's Paris meeting
> July 31 through August 5.
>
> On Tuesday, a collection of e-mail analysts, providers, and users will
> meet in New York for the Email Authentication Implementation Summit
> 2005, where Sender ID, SPF, DomainKeys, and DKIM will be put under the
> microscope. Microsoft, for instance, will be presenting results of its
> six-month use of Sender ID, while Bank of America will talk about
> authentication in online banking.
>
> In other e-mail and spam news, MX Logic said that in June zombies
> accounted for a record 62 percent of all spam. In comparison, May's
> tally was 55 percent, and April's 44 percent. "The continued
> proliferation of zombie PCs has levied a heavy cost on ISPs and end
> users," said Chasin. "Compromised PCs have resulted in millions of users
> being unknowingly blacklisted, often through no fault of their own."
>
> And the once-vaunted CAN-SPAM Act continues to be ignored by spammers,
> said Chasin, who noted that just 4 percent of all unsolicited mail in
> the first half of 2005 complied with the federal legislation. Since
> CAN-SPAM went into effect in January, 2004, compliance has averaged
> around 3 percent, with a high of 7 percent in December.
>
>
>
-------------------------------------
You are subscribed as roessler@xxxxxxxxxxxxxxxxxx
To manage your subscription, go to
http://v2.listbox.com/member/?listname=ip
Archives at: http://www.interesting-people.org/archives/interesting-people/