New Claims Broker working group at Identity Commons

[Paul Trevithick]

Here is the initial charter (v 0.4):

http://wiki.idcommons.net/Claims_Broker_Charter

If you are interested to be part of this working group, add your name to the wiki page above. Or email me.

This WG focused on use cases where we need to convey verified claims to the RP without the claims provider being able to know the identity of the RP and thereby track the user. You can do this sort of thing today in the real world: you can prove that you're over 21 to the liquor store owner using a drivers license without the drivers license issuer (claims provider) knowing that you've done so. 

As you can see the focus of this group is not about login. It assumes that the user logs in somehow to the RP site. And it assumes that the user logs in somehow to the claims provider (the IdP-like thing). 

You could think of this WG as a place to continue work on some of the best characteristics of Information Card technology, but in a way that complements existing authentication protocols, that doesn't require an active client, and that focuses on a narrow set of use cases which the proposed approach can address in a clear and definitive way.

BTW, if there's a better name than Claims Broker for the intermediary service, please suggest it.

Cheers,

--Paul

[Bob Pinheiro]

Since active clients are better than claims brokers in terms of user experience, security, and privacy (as stated in the Charter), active clients would be preferable for high assurance claims, whereas claims brokers may still be adequate for low assurance claims.  So for someone who desires to use an active client for high assurance claims, what would be the benefit to using a separate claims broker for low assurance claims?  Why not just use the same active client for both?  [Unless people are just going to use Infocards and OpenIDs for low assurance claims, then maybe active clients would be too much of a bother..........]

--
General mailing list for Information Cards, the Information Card Foundation, and open identity infrastructure. For more information, visit http://informationcard.net
To post to this group, send email to icf-co...@googlegroups.com
To unsubscribe from this group, send email to
icf-communit...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/icf-community?hl=en

 

[Paul Trevithick]

On Dec 15, 2010, at 4:42 PM, Bob Pinheiro wrote:

Since active clients are better than claims brokers in terms of user experience, security, and privacy (as stated in the Charter), active clients would be preferable for high assurance claims, whereas claims brokers may still be adequate for low assurance claims.  So for someone who desires to use an active client for high assurance claims, what would be the benefit to using a separate claims broker for low assurance claims? 

You can also use the claims broker service for high assurance claims. The active client merely improves the situation from a privacy (un-traceability), anti-phishing, and user experience point of view. In other words for a given LOA you may have to have to present the user with more authentication challenges if there is no active client. For example, you may have to send an SMS challenge to the user's phone, etc.

Why not just use the same active client for both?  [Unless people are just going to use Infocards and OpenIDs for low assurance claims, then maybe active clients would be too much of a bother..........]

Active clients can do both. But we've learned the hard way that active clients need to optional. 

[Anthony Nadalin]

> This WG focused on use cases where we need to convey verified claims to the RP without the claims provider being able to know the identity of the RP and thereby track the user. You can do this sort of thing today in the real world: you can prove that you're over 21 to the liquor store owner using a drivers license without the drivers license issuer (claims provider) knowing that you've done so.

 

So is the real issue collusion between the Department Of Licenses and liquor store you are trying to prevent? Or is that you are advocating using generated secondary claims from the Department Of Licenses to prove you’re over 21 w/o revealing additional claims from the Department Of Licenses?

--

[michael versace]

I'm in.

Added a requirement for the claims broker / claims provider to be able to provide current or past claims.  For example, to validate a claim of membership from a prior time.

--
General mailing list for Information Cards, the Information Card Foundation, and open identity infrastructure. For more information, visit http://informationcard.net
To post to this group, send email to icf-co...@googlegroups.com
To unsubscribe from this group, send email to
icf-communit...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/icf-community?hl=en

--
_______
Mike Versace

@versace57

[Paul Trevithick]

Just noticed that I didn't respond to this.

On Dec 20, 2010, at 7:57 PM, Anthony Nadalin wrote:

> This WG focused on use cases where we need to convey verified claims to the RP without the claims provider being able to know the identity of the RP and thereby track the user. You can do this sort of thing today in the real world: you can prove that you're over 21 to the liquor store owner using a drivers license without the drivers license issuer (claims provider) knowing that you've done so.

 

So is the real issue collusion between the Department Of Licenses and liquor store you are trying to prevent?

I'm trying to prevent the Dept. of Licenses from learning things about the person that you could argue they have no right to know.

Or is that you are advocating using generated secondary claims from the Department Of Licenses to prove you’re over 21 w/o revealing additional claims from the Department Of Licenses?

In addition to reducing linkability, yes, we'd like to follow the minimization principle that lies behind what you wrote. 

And, preaching to the choir, generated secondary claims (e.g. zero knowledge proofs) are a way to achieve both objectives. I personally don't know of any other way. 

 

From: icf-co...@googlegroups.com [mailto:icf-co...@googlegroups.com] On Behalf Of Paul Trevithick
Sent: Wednesday, December 15, 2010 11:21 AM
To: icf-co...@googlegroups.com
Cc: Harry Halpin
Subject: [ICF-Community] New Claims Broker working group at Identity Commons

 

Here is the initial charter (v 0.4):

http://wiki.idcommons.net/Claims_Broker_Charter

 

If you are interested to be part of this working group, add your name to the wiki page above. Or email me.

 

This WG focused on use cases where we need to convey verified claims to the RP without the claims provider being able to know the identity of the RP and thereby track the user. You can do this sort of thing today in the real world: you can prove that you're over 21 to the liquor store owner using a drivers license without the drivers license issuer (claims provider) knowing that you've done so. 

 

As you can see the focus of this group is not about login. It assumes that the user logs in somehow to the RP site. And it assumes that the user logs in somehow to the claims provider (the IdP-like thing). 

 

You could think of this WG as a place to continue work on some of the best characteristics of Information Card technology, but in a way that complements existing authentication protocols, that doesn't require an active client, and that focuses on a narrow set of use cases which the proposed approach can address in a clear and definitive way.

 

BTW, if there's a better name than Claims Broker for the intermediary service, please suggest it.

 

Cheers,

--Paul

 

 

-- 
General mailing list for Information Cards, the Information Card Foundation, and open identity infrastructure. For more information, visit http://informationcard.net
To post to this group, send email to icf-co...@googlegroups.com
To unsubscribe from this group, send email to
icf-communit...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/icf-community?hl=en

-- 

General mailing list for Information Cards, the Information Card Foundation, and open identity infrastructure. For more information, visithttp://informationcard.net

[Bob Pinheiro]

So while a cloud-based claims broker/agent can be used for both high and low assurance claims, the use of a claims agent for high assurance claims still seems to result in an inferior user experience as compared to an active client, since the user will need to jump through extra hoops (additional authentication challenges, etc) to achieve the same level of security.  Active clients still seem to provide an advantage in terms of privacy (un-traceability), anti-phishing, etc.  Another advantage of an active client that could store claims locally (maybe as u-prove tokens) is that these claims would still be available even if a cloud-based claims agent becomes unavailable for some reason. 

The motivation for cloud-based claims agents seems to rest on the assumption that active clients need to be optional.  I'm guessing that the main rationale for this is that users need the ability to access and manage claims from multiple devices, and device-specific active clients would require some mechanism to synchronize the claims across these different devices.  A cloud-based claims agent would allow these claims to be accessible from multiple devices.  A second rationale might be that active clients on user devices introduces other undesirable complexities that could be avoided through the use of a cloud-based claims agent.

While I don't deny these issues, I think that what should be primary in determining a path forward is to make sure that the user experience is as positive as possible, in terms of ease-of-use and other usability factors, security, and privacy protections.  This is especially important for consumers (as opposed to enterprise users), since I believe a key challenge for the identity community is to create an identity infrastructure/ecosystem that can enable greater acceptance and use of high assurance claims by consumers as well as service providers such as financial institutions, for the prevention of identity-related fraud.  I hope that the various identity organizations (ICF, Identity Commons, OIX, OIDF, Kantara) will devote sufficient resources to helping ensure that identity solutions provide the best possible user experience, especially as it relates to high assurance consumer claims.

Bob Pinheiro
Chair, Kantara Consumer Identity WG

[Craig Wittenberg]

[adding IC commons stewards]

 

I agree that we should “make sure that the user experience is as positive as possible, in terms of ease-of-use and other usability factors, security, and privacy protections.”  The more we can work together on this the better.

 

Hopefully those who are interested can discuss this on the 10^th .  I’ll propose a time.

 

Craig.