Here’s my starting point:

1. For a digital ID system to be accepted and genuinely useful it needs to grow, bottom up, from some perceived need.

2. Currently we’re not hearing a great yearning for digID coming from the grassroots. Rather, schemes are being imposed from the top down.

3. We have already evolved the digID mechanisms required to do what what currently want to do, everything from username/pwd logins to biometric scans at secure government facilities.

The counter that instituting strong digID will enable innovation doesn’t persuade me since it’s asking us to trade unknown benefits for known restrictions: I haven’t heard any ideas about the innovations other than the enabling of truly intrusive “digital rights management.”

But that’s not what I want to say here. This is:

If our current digIDs represent acceptable solutions to real problems — because if they weren’t, we wouldn’t have accepted them — then what does that tell us about the shape of a digID system that provides one-stop identification? What do we learn from our experience? How about:

– As a one-stop ID system, it lets us enter and update our ID info in a single form. That’s the improvement over today’s chaos.

– We own the info. No one has access to any of it without our permission. Ever. (The federated model is the most reassuring of all the one-stop schemes I’ve heard.)

– We get to establish profiles for various situations, so we can be on an ecommerce site and press a “Business Purchase” profile button that gives the appropriate credit card and address. and we can be on a game site and press our “Fun” profile button that loads in our Quake name.

– There is an identity firewall: You are not able or allowed to go from my online purchase of “Geeks with Guns” to my real world identity (unless, of course, I’ve allowed you to).

– No profile has default significance. That is, no one of the IDs I store in the digID system represents the “real me.”

– There is no ID overkill. Businesses and governments have an interest in over-specifying the digID system, establishing authentication far beyond what’s required for simple transactions (e.g., credit card number) because they want to control our purchases after the fact (= DRM) and control our ideas before the act (= Carnivore and TIA). The digID system we want, based on our experience so far, does not permit this type of unreasonable security. There is no user demand for it.

Have I just described PingID or some other scheme?

---

Eric responds on his blog site and Bryan Field-Elliot responds in email by pointing to an article he wrote 8 months ago on this topic.

Categories: Uncategorized dw