Recently, researchers at the University of Utah and the Department of Homeland Security announced the development of computerized lie-detection systems based on digitized studies of eye movements. Although these systems have not yet reached the practical applications stage, one can easily imagine the potentially profound benefits (and risks) of such technology.
Suppose, for example, that an airline or other security-conscious organization wished to avoid terrorist risks by screening passengers (or company visitors) via a questionnaire, coupled with observation of the subject, to detect signs of lying. This form of probing interview essentially aims to duplicate the security interviews that employees of El Al (the Israeli airline) routinely conduct. The difference is that such interviews could be conducted on a mass basis, even before the subjects arrive at the airport (or corporate location), by use of Webcams (essentially, as part of the online check-in process). Those who could not accomplish the remote check-in would be subject to in-person screening at the physical location.
Such a system might offer several advantages. Most significantly, the depth and effectiveness of security might be greatly enhanced. Instead of relying solely on physical inspection (such as metal detectors and x-rays), a questionnaire and lie detector might identify problematic passengers and visitors, even if they are not carrying obvious weapons. The online nature of the inquiry, moreover, could greatly speed the process of security clearance and allow passengers (visitors) to conduct the process at their leisure. The digital nature of the process would help avoid human error (due to fatigue or inattention) and eliminate any cultural bias in applying harsher security measures to particular ethnic, religious, or other groups.
The downsides include cost, effectiveness, and privacy. Development of a practical security system will no doubt require years of additional research, training, and implementation. The effectiveness of such systems, moreover, will require extensive testing. Studies of conventional polygraph testing generally show reliability in the 85 percent range, but results can vary, depending on the form of questioning, the skill of the examiner, and the degree to which the subject has learned to “game” the system with physical, mental, or pharmaceutical measures. Similar problems, no doubt, will arise with eye-tracking methods of lie detection.
Finally, these new systems might create significant privacy problems. The forms of questionnaires, the storage of data from the tests, and the procedure for following up on any security alerts triggered by the testing (including the expungement of “false positive” results) might all draw serious judicial and political attention from privacy advocates.
[Disclosure: The author is a partner in the New York City offices of Jones Day and chair of the firm's e-discovery committee. He teaches Electronic Discovery at New York Law School and Conflicts of Law at Hofstra Law School. The views expressed are solely those of the author, and should not be attributed to the author’s firm or its clients.]
— Steven C. Bennett is a partner in the New York City offices of international law firm Jones Day.