In recent years, reports requested by Congress from think tanks such as the Congressional Research Service as well as numerous attempts to pass legislation concerning data mining make it obvious that the government recognizes how powerful the technique can be. Examining the statements and actions of elected officials in addition to their inquiries concerning this subject demonstrates some of the important security, ethical, and legal issues that data mining produces in relation to Homeland Security.
Homeland Security has been a conceptual component of defense for centuries, but the specific term with its current implications was solidified during the reorganization of several United States government agencies following the terrorist attacks of September 11, 2001. The policies and goals of the recently formed Department of Homeland Security are constantly evaluated and revised, however the primary concern of the agency is to anticipate and prevent future terrorist attacks. In the wake of the September 11 attacks and subsequent military efforts to eliminate terrorist organizations around the world, terrorism has come to be regarded as the primary security issue opposing Homeland Security.
The use of data mining and decision support tools offers significant guidance in the effort to address the issue of terrorism. Knowledge gained through the use of data analysis might otherwise never been considered, due to the vast amount of data that are examined. Some factors that mitigate the potential value of data mining include the quality or quantity of data, validity of patterns revealed, and poor interpretation or analysis by insufficiently skilled personnel. These limitations, along with other problems, contribute to ethical and legal concerns that have been raised by critics who question and challenge the use of data mining on the grounds of those issues.
Ethical concerns facing Homeland Security are virtually inescapable. Balancing the need for security with the desire for privacy is an endeavor that cannot hope to satisfy everyone. The primary problem voiced by critics of programs such as the Total Information Awareness Project and the Computer-Assisted Passenger Prescreening System relates to "mission creep". This potential for data mining initiatives to be used for purposes beyond their stated intentions has been cited by opponents of such programs, and in some cases their arguments have led to legislation that resulted in the discontinuation of the offending projects.
Aside from the darker implications of using data collection and analysis for unauthorized objectives, attempting to obtain information that was never intended to be learned from particular data is a mistake that leads to inaccurate results. The unpleasantness of the idea that data mining applications intended to protect the population are being used against the citizens supposedly protected is further compounded by the huge potential for inaccuracies stemming from investigating the wrong problems. Civil liberties organizations seem less focused on such inaccuracies and more intent on the breach of privacy, however. Data mining based decision support systems used to identify and prevent terrorist activities unfortunately require vast amounts of personal data to analyze. Detractors do not feel the privacy sacrificed for these efforts is a fair trade for the claims of protection offered by supporters.
If data mining analysis never produced viable intelligence in the struggle to safeguard the nation from terrorism, the argument might be more one-sided. This is not the case, though, and advocates can point to the widely publicized allegations that one of the September 11 hijackers was identified long before the day of the attacks. The classified project Able Danger supposedly revealed Mohammed Atta to be a terrorist, and although testimony was never given by the project members who claimed he had been identified, their recollections were later backed up by others who worked on the project. If that information had been shared and resulted in the prevention of one or more of the attacks, it is quite possible that public opinion regarding the ethics of data mining would be more tolerant today.
Concerns about ethics have a direct impact on the legal issues connected with Homeland Security. The response to the September 11 attacks was a shift in policy from deterrence of terrorism to preemption using all available means to shut down the Al Qaeda network. This campaign included wide-ranging use of data mining and analysis to target members of the network that was almost unanimously approved in the aftermath of the attacks. As time went on, however, privacy concerns gained prominence and began to influence legislation aimed at curbing data mining efforts. Operating within the restrictions imposed by laws and regulations that limit the use of private information is a significant legal issue facing Homeland Security.
Although data mining activities have not been prevented by Congressional actions, numerous attempts have been made to halt or limit them. During the 108th, 109th, and 110th Congresses more than a dozen pieces of legislation were introduced designed to impose moratoriums or demand detailed reports about costs and effects on privacy. Senators Wyden and Feingold were especially aggressive towards the practice, and introduced at least eight amendments or acts with the intention of limiting data mining in some way between the two of them during the three Congresses. Many of the proposals were unsuccessful, but in 2003 amendments were passed that blocked funding for the TIA program. The controversial DARPA initiative was not dismantled entirely, but the legal restraints forced it to be restructured and the outcry contributed to the resignation of John Poindexter, the project overseer.
It is clear from the political opposition data mining faces that legal factors shape its effectiveness in Homeland Security. These legal factors are almost universally tied to the ethical issues of privacy rights. The tug-of-war between privacy and national security will probably not be won anytime soon, but it seems like a safe assumption that a new attack would shift the balance back toward tolerance of intrusion, and otherwise resistance to mining personal data for Homeland Security will continue.