Message-ID: <022901c562ee$0126ddb0$0e3eac18@MLANDE>
Date: Fri May 27 19:54:59 2005
From: mlande at bellsouth.net (Mary Landesman)
Subject: Bank of America SiteKeys ineffective?

Thanks for the clarification.

Presumably then, the attackers can't just take the harvested details and
login to the accounts; they would have to introduce a next level, i.e. the
man-in-the-middle attack described below?

I agree that it's not foolproof, but I would think it can help alleviate at
least some phishing scams - specifically those that just try to harvest a
large number of account details by enticing users to login to fake websites.
Stopping some is better than nothing, so long as BoA (and others) view this
as one step out of many and not as a silver bullet.

> Many people regularly dump their cookies

That's an interesting point. I encourage people not to use cookies to
remember their bank login info - particularly those using laptops. Wonder
how many will get frustrated by the constant challenge/response, allow the
cookie and remembered login, and then have their PC accessed or stolen? I
hope they will institute two sets of cookies; one just to remember the
machine and a second for the login. This way, you can bypass the more
extensive c/r, but still have to supply the standard login credentials to
access the account.

> training 13 miillion customers how to
> inspect their certificates and actually have people look at their
> certificates is also probably unrealistic.

Yeah. The yellow lock - simple as it seems in principle - is a good example
of that problem.

-- Mary

----- Original Message ----- 
From: "Mike N" <niceman@....net>
To: <full-disclosure@...ts.grok.org.uk>
Sent: Friday, May 27, 2005 1:39 PM
Subject: Re: [Full-disclosure] Bank of America SiteKeys ineffective?


>From: "Mary Landesman" <mlande@...lsouth.net>
>Subject: Re: [Full-disclosure] Bank of America SiteKeys ineffective?

> >From my read of the news.com article and admittedly limited knowledge of
> SiteKeys, it does not seem to me their intent is to make sure the user
> knows  they are at a legitimate BOA page. Rather, it seems to me
> the intent is to
> ensure that if Betty Boop logs into her BOA account, that she's doing so
> from a pre-authorized Betty Boop specified computer.


I found the official press release at

http://www.bankofamerica.com/newsroom/press/press.cfm?PressID=press.20050526.03.htm

   In the press release, one of the 2 key goals is to "Confirm the Web
site's validity."  From the description, it will do no such thing - it only
confirms a possible link from their browser to the BofA web site, not that
they are linked correctly and solely to the proper BofA web site.

  Even the challenge-response scenario is nearly useless.   If for some
reason the  phisher in the middle couldn't steal the secure cookie and pass
it on to the real site, the customer might fall for the challenge-response
questions being relayed from the phisher and answer them; the phisher would
end up with the challenge-response answer as well as the login.   Many
people regularly dump their cookies for privacy reasons; those people will
become used to seeing the challenge-response and they won't realize they're
being taken.

 The press release mentions that they are using PassMark
http://www.passmarksecurity.com .

   The PassMark is better than nothing, but doesn't accomplish anything in
the end except to make the customer feel better.   It's not as effective as
inspecting the HTTPS certificate, but training 13 miillion customers how to
inspect their certificates and actually have people look at their
certificates is also probably unrealistic.




_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/

Powered by blists - more mailing lists