ISIS Blogs

(edit: I have been known to be crazy in the past, this may have been one of those times)

Nitesh recently gave a presentation on his concerns that the privacy of our personal information and the privacy of the data that we generate on the web (google search for example) is not being adequately protected and that he has been spending a small amount of time researching technical innovations which could help balance this deficiency.

If you are in that small minority of people who know who data brokers are, who know how private investigators are able to perform their jobs, or who know the exact wordings and loopholes and the laws supposedly protecting our privacy, then you are no doubt worried that the abstract feeling you call “privacy” may have gotten left behind in the 20th century, never to return. If you’re one of the people in our ISIS lab like Nitesh, you’re probably dying to come up with a brilliant solution that will solve this problem like every other that you come across and conquer.

But is it possible to use technology to solve the existence of an industry or solve the already widely spread personal information flowing through companies databases?

I would argue that no, it is not possible, and if it is possible, that it’s one of the hardest technological issues there is outside of NP. Companies that catalogue your information already exist. They know who you are, what you eat, who you’re married to, where you went to school, and your sexual orientation. Short of creating a massive worm that infects every computer in the world and destroys all the PII (personally identifying information) it finds, there is no way to take that back. It’s out there and there is no piece of code you can write, no math you can dream of, that will tip the balance of power back into the individual’s hands.

Therefore, it appears to me that this problem is a legal one. The only entity with the authority to effect such change is our government, who after making changes to its laws can declare illegal the actions we don’t like. While it’s not our job in this case to come up with a purely technical solution to the problem of Web Privacy, we can help come up with an infrastructure to enable a particular formation of laws and regulations to store personal information in a way to protect privacy and demonstrate to other lawmakers and industry that the system can work.

One such person who came up with an infrastructure for managing the privacy of personal information was a graduate of the ISIS Lab and a good friend of mine, Michael Aiello. Aiello’s idea, in its simplest form, is to:

• Centralize the storage of any personally identifying information which can be used to directly locate or communicate with you in a government agency (let’s call it the IA - the Information Administration).
• Provide you the ability to manage your own data at the IA by only distributing information to those companies which you approve and by allowing you to modify and correct information about yourself.
• Force outside entities to query the IA for your identifying data by making the storage of such PII illegal.

You can think of it as Google Checkout for your PII. Every time someone needs your credit card, they have to go to Google and request it from them. You then allow or cancel the charge and then the requestor gets access to your money for a period of time. You can log in to Google at any time and change your billing address and credit card in one place and not have to worry about everyone else using the old one. The entities you are doing business with no longer store your credit card number themselves.

Without going into further detail, I’d like to run through some of the scenarios he outlines to demonstrate this idea’s utility.

Scenario 1:
I receive unsolicited mail from Company X.

By sending me mail, Company X knows at least my name and address and possibly whether I would be interested in their mailing based on my purchase history. I go online to the IA website, login to my account, and look through the history of accesses to my personal information. Company X is not in the list and therefore, this was an illegal act on the part of Company X.

Scenario 2:
A company is hacked by an attacker who obtains all of the social security numbers of their customers.

By making the storage of certain personally identifying information illegal, the attacker should only have access to a list of unidentified social security numbers without names, addresses, or other information to accompany them. If the attacker wants to use a social security number, he has to query the IA for identifying information (because no one else has it). This query will show up in your access logs and you will be granted the authority to approve or deny it.

To prevent abuse of the IA, a small fee for a high number of queries will be charged. The profits from this fee will be the agencies only funding, helping to force the IA to remain effective and efficient instead of relying on handouts from (and becoming beholden to) Congress.

Without going any further, what are your thoughts on this idea? Say this system were implemented and laws were enacted to enforce it, how would identity theft and other privacy violations occur despite it? Can you identify any deal breakers that would prevent this system from working?

In part 2 of this series on privacy I’ll discuss how Aiello’s idea is similar to that of Daniel J. Solove in his paper “A Model Regime for Privacy Protection.” Solove is a premier researcher of privacy with a specialty in issues related to technology. Solove collaborated with EPIC in writing his paper and goes into detail on the exact changes to current laws necessary to make his system work.

Stay tuned…

-Dan
dguido@gmail.com

btw, first post! :-)