[DG-IDPro] Attributes and Context [WAS: IdM Poster.]

Disclaimer to start: my contribution to this thread is intended to not judge correctness or incorrectness of anyone's views. The poster and dialogue sparked an interesting line of thinking. "Under what circumstances would identity data and authentication factor data be considered to be identity attributes" An answer took a while because I deal with IDP, CSP, AA, RP terms all day. I think one answer is: when you are an external observer or data analyst looking at a data stream and have no direct knowledge of any attribute databases or individual entities. or: Data analytics on data streams where the observer is not one of IDP, CSP, AA or RP. The mode that I typically work in is where the individual entity is asserting some claim. Or as a Verifier evaluating an authentication assertion. Or as an attribute authority with a database of well-organized data. When you take those parties out of the observation, then everything on the wire could be: infrastructure management data, transactional data, other metadata, and possibly attributes that could be ascribed to an individual entity. I also see that Nat's observation of the ISO definitions is compatible with my musings. andrew. *Andrew Hughes *CISM CISSP Independent Consultant *In Turn Information Management Consulting* o +1 650.209.7542 m +1 250.888.9474 1249 Palmer Road, Victoria, BC V8P 2H8 AndrewHughes3000@gmail.com ca.linkedin.com/pub/andrew-hughes/a/58/682/ *Identity Management | IT Governance | Information Security * On Tue, Mar 7, 2017 at 6:15 AM, Natale, Bob wrote:

That’s true … but not necessarily the total dichotomy implied by the assertion that “Identity is NOT Authentication” … authenticator attributes can also be identifier attributes and vice versa … the closer an authenticator attribute (like a human fingerprint) is to a unique identifier (like a human fingerprint paired with some threshold set of appropriate biographic, biometric, behavioral, and social attributes), the murkier the line between identity and authentication might become.

Or so it seems to me…. Things might vary a bit, with respect to what is a valid identity attribute, between human person identities and virtual entity identifiers as well … but that’s probably a topic for a different thread!

Ultimately, however, I strongly favor keeping the concepts of identity, authentication, authorization, and access control distinctly defined, as that is what holds in the general case.

Avanti,

BobN

*From:* dg-idpro-bounces@kantarainitiative.org [mailto:dg-idpro-bounces@ kantarainitiative.org] *On Behalf Of *Jim Willeke *Sent:* Tuesday, March 07, 2017 6:26 AM

*To:* dg-idpro@kantarainitiative.org *Subject:* Re: [DG-IDPro] IdM Poster. (thats wrong)

As always, statements without context are philosophical discussions.

Identity is NOT Authentication.

Authentication is "things that RPs need to know about Subjects in order to [help] decide whether or not to accept a message, document etc."

Authentication is the process of establishing confidence in the Identification of an Entity. That is confidence that the Identification is authentic.

Levels of Assurance or Vectors of Assurance address the degree of confidence in an assertion of Authentication.

--

-jim Jim Willeke

On Tue, Mar 7, 2017 at 5:02 AM, wrote:

In essence, I think most IDAM professionals would agree that attributes are things that RPs need to know about Subjects in order to [help] decide whether or not to accept a message, document etc. Some of the nice questions we're all dealing with currently are:

- are attributes (ie what someone is) more important than "identity" (ie who someone is)?

- how do you know that a given attribute about a Subject is true of the Subject?

- that is, what authority vouches for the attribute?

- and how do you know that a presented attribute is bound to the Subject and isn't being replayed?

If an attribute is something that we need to know about someone, then clearly passwords are something else. Likewise for PINs (the cool thing about PINs when at matched on-card is that nobody other the Subject ever knows the PIN). And CVVs.

And then there is biometrics. There are broadly two modes of biometric presentation: One-to-One, where it is generally preferred that the biometric is matched locally in order to unlock a device (ala FIDO, or Apple iTouch), and One-to-Many (often tellingly called "identification") where I suppose the attribute could be regarded as an attribute. But the general aversion to One-to-Many matching of biometrics points to an ideal where biometrics are NOT identity attributes!

Cheers,

Steve.

Stephen Wilson

*Lockstep Group *

W: http://lockstep.com.au

T: @steve_lockstep

*Lockstep Consulting provides independent specialist advice and analysis *

*on digital identity and privacy. Lockstep Technologies develops unique *

*new smart ID solutions that enhance privacy and prevent identity theft. *

-----Original Message----- From: "David Chadwick" Sent: Tuesday, 7 March, 2017 6:07pm To: dg-idpro@kantarainitiative.org Subject: Re: [DG-IDPro] IdM Poster. (thats wrong)

Hi Kaliya

Glad you are not in my class!

Seriously though, passwords are identity attributes if one regards every piece of information that is associated with a user as an identity attribute. But they are clearly not identifiers in the general case, as they do not uniquely identify anyone, given that 'password' and '123456789' are two of the most common passwords on the Internet. However, if you have a very strong password then it is possible that it could be an identifier, if you are the only person in the world using that password.

regards

David

On 07/03/2017 04:24, Kaliya Identity Woman wrote:

...

HI ID Pro's

As those of you know who attended the ID-Pro breakfast at RSA.. I'm in the new Masters of Science in Identity Management and Security at UT Austin.

There have been some challenges in what has been taught... including that the factors of authentication are not that...but "identifying Information" or as in the poster below says "Identity Attributes"

They also have taught that password are identifiers (yes this was actually taught)... in this poster on the other side they are identity attributes..yes identity attributes. Sigh. I have raised issues about these two things that have been taught...and well not gotten very far. (besides being told i'm a "bad student" and "unwilling to learn".

But now they have this fabulous poster. I'm hoping some of you with blogs or twitter handles can point at the poster - references it and explain why both things are wrong. (cause they, specifically Dr. Barber and Dr. Doty don't believe me.

Or maybe this group could write a joint letter explaining its 'wrongness" it snot great that this center is putting out this information...it doesn't help us in the long run get explaining this stuff right.

Here is the post on their site with the poster. https://identity.utexas.edu/infographics/identity- attributes-and-the-identity-ecosystem

Here is Dr Barbers faculty page - http://www.ece.utexas.edu//people/faculty/suzanne-barber http://www.ece.utexas.edu/people/faculty/suzanne-barber

Dr. Doty's

https://www.ischool.utexas.edu/people/person_details?PersonID=22

_______________________________________________ DG-IDPro mailing list DG-IDPro@kantarainitiative.org http://kantarainitiative.org/mailman/listinfo/dg-idpro

_______________________________________________ DG-IDPro mailing list DG-IDPro@kantarainitiative.org http://kantarainitiative.org/mailman/listinfo/dg-idpro

_______________________________________________ DG-IDPro mailing list DG-IDPro@kantarainitiative.org http://kantarainitiative.org/mailman/listinfo/dg-idpro

_______________________________________________ DG-IDPro mailing list DG-IDPro@kantarainitiative.org http://kantarainitiative.org/mailman/listinfo/dg-idpro