[readingaloud]

I'm attracted to the Kindle, but I'm really alarmed at the idea that the device is going to report back to Amazon, apparently in considerable detail, about what I'm doing with it. Amazon already knows a lot about what books I favor (at least the ones I've bought from them), but the Kindle apparently is going to tell them how much time I spend reading them, where I put bookmarks, what notes I make--all sorts of things that I find really intrusive.

Now, I can see why Amazon wants at least some of this information. And, up to a point, I'm OK with it--if Amazon can make more nuanced recommendations based on knowing more about how and what I read, we both benefit. But the rights they're claiming are breathtakingly broad, and I think we should all be worried about what they might some day decided to do with this information that they're accumulating.

The only gesture Amazon has made in the direction of reassuring us that the information will not be abused is to say that it's subject to their privacy policy. But, if you read that policy, you'll see that it's not very limiting. It says, for example, that they're not in the business of reselling customer information, but not that they won't decide to get into that business at some point. And the privacy policy can be changed whenever Amazon wants to change it.

I think we need something more from Amazon. Let me suggest that they commit themselves, now and in the future, to some sensible limitations to how they use the information harvested from Kindle:

(1) Amazon should have the right to use and distribute statistical aggregatations of the data for any purpose, but this right should be strictly exclude any aggregations small enough to risk the identification of individual users.

(2) Amazon should have the right to use individual data any way they like for the purpose of enabling the algorithms that suggest additional purchases. But this right should specifically forbid them from making the reading history of an identifyable individual customer available to any person other than the customer himself or herself. With the approval of the customer, this information can also be displayed to a customer-service representative working on a service request initiated by that customer, and Amazon can even refuse, when necessary, to provide certain sorts of customer service to those who do not consent.

I think these two rules strike a reasonable balance between Amazon's legitimate interest and the reasonable privacy rights of the customer.

Further, I think that Amazon should structure it's databases so that these limitations will be, to the greatest possible extent, self-executing. By this I mean that I want Amazon to be able to say, truthfully, that it's simply not possible for them to honor information requests that violate these strictures.

I'd like, for example, Amazon to be able to tell the Chinese government that they simply cannot provide a list of the reading materials used by a Chinese expat living in the U.S. I don't want to rely on their assurance that they would not honor such a request, because the day may come when they have to make a choice between honoring such a request, or losing access to the huge Chinese market. It's better for all concerned, including Amazon, if they're able to say that providing this kind of information is not just against company policy, but that it simply can't be obtained. And I want Amazon to be able to give the same answer when the request comes from the U.S. Justice Department.

Similarly, I want to be assured that no rogue Amazon employee will ever be able to snoop into my affairs--not just that they'll be told not to, but that it won't be possible for them to do it.

Now, I can imagine that I've overlooked some important element here--that I'm not allowing for some reasonable use of the data, or that there are other limitations that ought to be imposed. This is meant to start a conversation about the proper rights and limitations of Amazon with respect to the data.

--Readingaloud