Anne & Lynn Wheeler <[EMAIL PROTECTED]> writes:
> That very theory is what the MasterCard and Visa networks are
> supposed to be proving with SET, the Secure Electronic Transaction
> protocol, which requires digital certificates for banks, merchants,
> and cardholders. A purchaser would use the digital code, rather than
> a card number, to initiate a transaction; the certificate would
> represent authentication by a bank or other certificate authority.
> 
> SET has gained some acceptance overseas but almost none in the
> United States, therefore contributing little to the mainstreaming of
> digital certification.
> 
> The concept has, however, attained some level of business-world
> consciousness through initial public offerings this year by two
> specialists in the field, Verisign Inc. and Entrust Technologies
> Inc. These companies and others stand ready to give banks the
> software or outsourcing support they need to authenticate people
> doing business facelessly on the World Wide Web.

two issues from the historical news article posting
http://www.garlic.com/~lynn/2005l.html#34 More Phishing scams, still no SSL 
being used

is that their "light" digital certificates are also referred to has
relying-party-only certificates 
http://www.garlic.com/~lynn/subpubkey.html#rpo

... since rather than carrying the actual information, they just carry
some sort of index pointer into a business relationship management
infrastructure (like account number). the relationship management
infrastructure contains all the real information.

however, it is trivial to show that such "light" certificates are
redundant and superfluous when the business process has to access the
infrastructure containing the real information. this is also somewhat
a trivial logic operation when you take the original design point for
digital certificates is providing relying parties with information in
an offline environment when the relying parties had no other recourse
to the real information; aka therefor by definition, if the relying
parties have access to the real information ... the original purpose
and justification for digital certificates is invalidated.

the other issue is that even for "light" certificates the
infrastructure overhead for appending certificates ran 4k to 12k
bytes. when you are talking about the basic payment card
infrastructure where typical message size is 60-80 bytes, the appended
certificate paradigm represents an enormous payload bloat of one
hundred times (two orders of magnitude) ... for otherwise redundant
and superfluous certificates.

the basic technology is asymmetric key cryptography, where what one
key (of a key-pair) encodes, the other of the key-pair decodes.

a business process has been defined called public key ... where one of
the key-pair is made freely available and the other key is identified
as private and kept confidential and never divulged.

a further business process has been defined called digital signature 
... which represents "something you have" authentication ... from
3-factor authentication paradigm
http://www.garlic.com/~lynn/subpubkey.html#3factor

* something you have
* something you know
* something you are

digital signatures implies that some entity has access and (presumably
sole) use of a specific private key.

existing relationship management infrastructures can upgrade their
shared-secret based authentication
http://www.garlic.com/~lynn/subpubkey.html#secret

to digital signature, by registering a public key in lieu of pin,
password, ssn, date-of-birth, mother's maiden name, etc. while in
secret-based infrastructures, the same value is used to both originate
as well as authenticate. in public key scenario for digital signature,
the public key is only used to authenticate (and can't be used to
originate or impersonate).

from the PAIN security acronym

P ... privacy (or sometimes CAIN, confidential)
A ... authenticate
I ... integrity
N ... non-repudiation

it can be easily demonstrated that relationship management
infrastructures tend to have very high integrity reguirements
(regarding all details of the relationship as well as the
authentication information).

however, many business infrastructures that make heavy use of their
relationship management infrastructure for numerous business process
are also at risk of exposing the authentication information. when this
authentication information is public key, it tends to not be a big
deal. however, when the authenticaton material are secrets, then there
is an enormous privacy requirement (since obtaining the secrets also
enables impersonation, fraudulent transactions, account fraud, etc).

Using secret-based authentication can create enormous dynamically
opposing objectives for relationship management infrastructure ... on
one hand the relationship management infrastructure has to be readily
available in support of numerous business operations .... and on the
other hand, the secret-based privacy requirements are none but
extremely constrained business operations can access the information.

one such description is my old security proportional to risk posting
http://www.garlic.com/~lynn/2001h.html#63

and a whole host of postings on skimming and harvesting of
(secret-based) authentication material (that can be leveraged
to perform fraudulent transactions)
http://www.garlic.com/~lynn/subpubkey.html#harvest

as an aside ... there was a report within the past couple years that
something like 70 percent of identity/account fraud involved insiders.
there has been some additional, similar reports. there were a number
of these news URLs in the past couple days:

Bank workers biggest ID theft threat; Insiders with access to data may
pose 70% to 80% of risk
http://deseretnews.com/dn/view/0,1249,600145529,00.html

Banks Face Challenge In Screening Employees To Avert Inside ID Thefts
http://www.banktech.com/aml/showArticle.jhtml?articleID=164904297

-- 
Anne & Lynn Wheeler | http://www.garlic.com/~lynn/
_______________________________________________
mozilla-crypto mailing list
[email protected]
http://mail.mozilla.org/listinfo/mozilla-crypto

Reply via email to