Monday, November 19, 2007

Identity fragmentation

Francis Shanahan has been thinking about his digital identity.

"A few weeks ago I joined Facebook (after much resistence). Facebook sucks you in, making it so easy to give up bits of information about yourself, many times without even realizing it. It occurred to me that I'm leaving pieces of my identity everywhere.

Last night I took a stab at listing out the various entities that know me, regardless of how they know me. The list is overwhelming. It quickly became apparent that to develop a comprehensive list was not feasible. What I ended up with was a good all around representation. I then generalized it to include things not solely pertaining to me as an individual (e.g. I'm an immigrant, I can never have govt clearance).

With all the talk of identity and claims federation, this was a good way to step back and at least understand the problem space a little better. I'm sure there are other such diagrams out there but the benefit for me was to go through the process of drawing it rather than take one off the shelf.



[...]

In theory these entities could share Identity Providers. I believe we'll start to see this quite soon in the social networking space most likely through OpenSocial.

Ultimately, Identity Providers themselves will begin to exchange claims although it's questionable if this is an appropriate model.

This is by no means a complete model. I worry that I'll never be able to effectively manage all the pieces of me that I'm absent-mindedly handing out."

His diagram demonstrates, more clearly than I have seen anywhere else, the fragmented nature of online identity, the amount of personal information we thoughtlessly release to a variety of entities, and the potential power of aggregation when others (e.g. Google and DoubleClick) start linking this all together. For the skeptics on the privacy front, before you say "so what", factor fingerprints, iris scans and other poorly secured biometric data into the mix and think about the potential implications.

No comments: