Fast Company

Government Experimenting With Identity Technologies

Yesterday the United States Government in collaboration with industry announced a few pilot projects using emerging open identity technologies for citizens to use when interacting with government sites. This is part of the The Obama administration's open government memorandum calling for transparency participation, collaboration more real.

The Obama administration open government memorandum called for transparency participation, collaboration and federal agencies have begun to embrace Web 2.0 technologies like blogs, surveys, social networks, and video casts. Today there are over 500 government Web sites and about 1/3 of them require a user name and password. Users need to be able to register and save information and preferences on government Web sites the same way they do today with their favorite consumer sites, but without revealing any personally identifiable information to the government.

Yesterday the United States Government in collaboration with industry announced a few pilot projects using emerging open identity technologies for citizens to use when interacting with government sites. I use the word interacting very deliberately because the government doesn't want to know "who you are" and has gone great lengths to develop their implementations to prevent citizens from revealing personally identifiable information (name, date of birth etc).

How would you use this?--well imagine you are doing an in depth search on an NIH (National Institute of Health) Web site--and you went back to the site many times over several months. Wouldn't it be great if the site could "know" it was you and help you resume your search where you left off the last time. Not your name and where you live but just that you were there before.

The Identity Spectrum helps us to understand how it all fits together.

Spectrum of IDAnonymous Identity is on one end of the identity spectrum--basically you use an account or identifier every time go to a Web site--no persistence, no way to connect the search you did last week with the one you did this week.

Pseudonymous Identity is where over time you use the same account or identifier over and over again at a site. It usually means you don't reveal your common/real name or other information that would make you personally identifiable. You could use the same identifier at multiple sites thus creating a correlation between actions on one site and another.

Self-Asserted Identity is what is typical on the Web today. You are asked to share your name, date of birth, city of residence, mailing address etc. You fill in forms again and again. You can give "fake" information or true information about yourself--it is up to you.

Verified Identity is when there are claims about you that you have had verified by a third party. So for example if you are an employee of a company your employer could issue a claim that you were indeed an employee. You might have your bank verify for your address. etc.

The government pilot is focused on supporting citizens being able to have pseudonymous identities that function only at one Web site--the same citizen interacting with several different government Web sites needs to use a different identifier at each one so their activities across different government agencies do not have a correlation.

It is likely that some readers of this blog know about and understand typical OpenID. Almost all readers of this blog do have an openID whether they know it or not because almost all the major Web platforms/portals provide them to account holders--MySpace, Google, Yahoo!, AOL etc.

So how does this work with OpenID?

Typical OpenIDTypically when logging in with OpenID on the consumer Web you share your URL with the site you are logging into--they redirect you to where that is hosted on the Web--you authenticate (tell them your password for that account) and they re-direct you back to the site you were logging in. (see this slide show for a detailed flow of how this works). Using OpenID this way explicitly links your activities across multiple sites. For example when you use it to comment on a blog-- it is known your words come from you and are connected to your own blog.

Using the OpenID with Directed identity--de-links your the identifiers used across different sites but still lets you use the same account to login to multiple sites.

Directed IdentityWhen you go to login to a site you are asked to share not "your URL" but just the name of the site where your account is--Yahoo! or Google or MySpace etc. you are re-directed to that site and from within your account a "directed identity" is created--that is a unique ID just for that Web site. Thus you get the convenience of not having to manage multiple accounts with multiple passwords and you get to store preferences that might be shared across multiple ID's but you don't have identifiers that correlate--that are linked across the Web.

How does this work with Information Cards?

This is a complementary open standard to OpenID that has some sophisticated features that allow it to support verified identities along with pseudonymous & self asserted identities. It involves a client-side piece of software called a selector--which selector helps you manage your different identifiers using a card based metaphor, with each digital "card" representing a different one. Citizens can create their own cards OR get them from third parties that validate things about them.

The government is creating a privacy protecting "card profile" to be used in the pilot program. It is NOT issuing identities.

Trust Framework are needed to get it all to work together.

From the press release yesterday:

"It's good to see government taking a leadership role in moving identity technology forward. It's also good to see government working with experts from private sector and especially with the Information Card Foundation and the OpenID Foundation because identity is not a technical phenomenon--it's a social phenomenon. And technological support for identity requires the participation of a broad community and of representatives of government who define the legal framework within which identity will operate," said Bob Blakley, Vice President and Research Director, Identity and Privacy Strategies, Burton Group. "Today's announcement supplies the most important missing ingredient of the open identity infrastructure, mainly the trust framework. Without a trust framework it's impossible to know whether a received identity is reliable."

The OpenID Foundation and Information Card Foundation wrote a joint white paper to describe how they are working on developing this. From the abstract:

[They] are working with the U.S. General Services Administration to create open trust frameworks for their respective communities.
These frameworks, based on the model developed by the InCommon federation for higher education institutions, will enable government Web sites to accept identity credentials from academic, non-profit, and commercial identity providers that meet government standards. These standards are critical as they represent the government's resolution of the challenging and often competing issues of identity, security, and privacy assurance. Open trust frameworks not only pave the way for greater citizen involvement in government, but can enable even stronger security and privacy protections than those typically available offline.

These are all exciting developments but there is much more to do.

Looking (far) ahead there may be the opportunity to do selective disclosure--combining anonymity with verified identity.

How do these go together--you can take a verified identity claim say your birth date then using cryptography strip the specifics away and just have a claim that says you are "over 21". Then using an anonymous identifier you have selectively disclosed your age without giving away your date of birth.

You could imagine this would be handy for citizens wanting to communicate their opinions to their member of congress without revealing their actual name and address - they could "prove" using a verified claim they live in the district but not reveal who they are. This aspect of what is possible with the technology is VERY forward looking and will take many years to get there. There is enormous potential to evolve the Web with this emerging identity layer.

I would like to invite all of you interested in being involved/learning more to attend the Internet Identity Workshop in Mountain View California November 3-5. I have been facilitating this event since its inception in 2005. It is truly amazing to see how far things have progressed from when we were 75 idealistic technologist talking about big ideas. at the Hillside Club in Berkeley. It is also some what daunting to think about how much farther we have to go.


Kaliya Hamlin is the co-founder, co-producer and facilitator of the Internet Identity Workshop an open industry forum that has been meeting every 6 months since the fall of 2005. Known as the Identity Woman (the title of her blog) and is an independent expert in user-centric digital identity.  She was named by Fast Company as one of the most Influential Women in Technology.

Add New Comment