There are numerous technology and security concerns with any proposed "mega database." Here are a few that I'll discuss briefly in the following sections:
Developing means of bringing the right information together
Developing an effective method of querying the resulting large volume of information
Keeping both the data and the database secure
A Single Point of Disaster
Keeping this quantity of information about our nation and fellow citizens is a concern, as it will make a truly amazing target for hackers, terrorists, and enemy states. Just imagine what havoc someone could wreak by accessing all the information that the U.S. government holds on its citizens, residents, etc.
The potential scope for terrorist activities is so great that it's difficult to list any single set of possible consequences. Consider simple identity theft: using personal information to impersonate an unsuspecting individual, and then ordering bank transfers to launder or steal money, or just to purchase items on someone else's credit.
While standards certainly exist for transmitting and storing information securely, there are also potential security solutions for a great majority of today's well-known vulnerabilities. But these security precautions are often complicated and difficult to implement correctly. Further, they can be expensive, may lead to inefficiencies, or may not be compatible with a particular environment.
Keeping data safe will require a great deal of attention and a commitment to expend the resources required to implement and maintain this high required level of security. Further, there needs to be a commitment to secure the database not only when it's first being established, but throughout its lifetime.
How Do You Process "Mega Data"?
The point is to use this collection of data to predict and prevent events such as those of 9/11. This objective requires an organized and efficient database schema that will intelligently categorize all the information, allowing for quick and effective queries that can search and correlate all the disparate information. This poses a substantial database-management problem, but the technology already exists to make practical use of large databases and to train end users, and such technology can certainly be brought to bear on this project. However, solutions must be considered carefully for the database to aid in effective recordkeeping and decision making, and the cost of these solutions must be taken into account.
If We Build It, Who Will Come?
Who will access this database? It has always been said that government needs to share information so that all agencies can coordinate activities and defend our homeland. However, who exactly will have access to the database and be privy to the secrets it holds? What processes will they go through to be authorized to access, view, analyze, change, and potentially use this comprehensive and sensitive information? What stops someone who is authorized to use this information from abusing the privilege and selling it to the highest bidder? According to a recent CSI/FBI computer crime survey, the majority of cyber events still originate from within; users (employees, contractors, and so on) with legal access already have crossed numerous network defenses.
Further, what will be the ultimate number of people with this access? It's well understood that the wider the dissemination of information or accesswhether to a room, an area, a computer, or a databasethe harder it is to keep that information or access secure and private.
How Will They Come?
Where will the people who access the database be located? Within or outside the U.S.? Will access be restricted to domestic counter-terror officials and agents, or will agents in the field also be privy to the database? For example, will foreign-based CIA agents be able to access the database, although laws on IT matters outside U.S. borders are significantly different from our domestic laws?
Will these queries be run over the public Internet? (I hope not.) Will we have to develop a new global communications network, such as the GovNet concept being discussed by the President's Critical Infrastructure Protection Board? Or perhaps this will run over one of the Department of Defense communications networks?
Whether a new network has to be built or an existing private network will be used, the full length of this network needs to be protected, as simply tapping the communication lines can give intruders access to the queries that are conducted and the analysis that's performed. Further still, the network must be protected from any attempts to bring it down and make it crash, whether those attacks are physical (bombing of data centers) or logical (denial-of-service attacks).
Securing the Database Itself
We know that the database and the network upon which it resides will have to be secured. Among other things, this implies physical security by men with guns. Persons without clearance should not be able to get close to the database. In addition, access should be restricted to a small and trusted group of database administrators (DBAs); as mentioned before, the more people who have access, the harder it is to secure the information and keep it private. Restricting the DBAs will likely involve forcing them to gain access to the systems through multiple gates with both computer and human agents guarding the entrances. One option might be using a biometric smart card with photo ID access system to gain entry to the data center in which the database is housed. A guard can look at the photo ID and pass or reject someone attempting access, and a biometric card reader can scan the card, requesting a PIN and one or more of the many biometric scansfingerprint, palm scan, retinal scan, voice print, etc.
The facility and its equipment must also be protected from physical attack. We've already seen that some terrorists have an affinity for suicide bombing. It's not inconceivable that they would utilize this tactic against institutions housing the database or even locations from which the database could be accessed, such as office buildings with end-user workstations. We also don't want a situation where one of the "good guys" is looking at a computer screen of query results or the background of a particular person, and a foreign spy or agent of a terrorist organization is looking over his shoulder or peeping through binoculars from a window across the street to read the sensitive information. Nor do we want anyone in the vicinity to be able to access the electromagnetic signals that emanate from the computing devices used to access the database.
The equipment also must be protected from theft. A thief may be able to steal information still stored in the stolen machine's cache. A password may be found in the cache or on a file on the machine, essentially leaving the database open to compromise.
Environmental controls as well as an efficient backup power supply will have to be maintained. We can't have the system crashing and losing all its data.
These security measures can certainly be enacted. The technology and the military know-how already exist. But will any of these security measures somehow affect the legitimate activities of the groups using these facilities? And of course there will be a cost associated with these security measures, which must be taken into consideration.