Passenger screening gimmick stuck at the gate

'CAPPS-3' not ready for takeoff

The US Transportation Security Administration (TSA) is behind schedule in developing a new terrorist-busting database system called Secure Flight, a report by the Government Accounting Office (GAO) says.

After confronting the obvious defects in the old pre-9/11 CAPPS (computer-assisted passenger pre-screening system), which allowed 19 violent terrorists to board flights on a single morning, the TSA set out to develop CAPPS-2, supposedly an improvement. When that project failed to result in a working system, TSA announced that it would re-work the entire scheme.

Proposed improvements included letting the government, rather than airlines, administer the system, so that secret counter-terrorist intel could be used, and merging airline passenger data with commercial data such as that stored by privacy invasion outfits like ChoicePoint.

TSA got off to a strong start, successfully changing the system's name, for example, but has since fallen behind on lower-priority modifications, such as establishing privacy standards, and basically making it work.

Congress established ten milestones that Secure Flight must pass before its intended roll-out in August. Of these, nine remain to be satisfied. (An advisory committee has actually been chosen as required, but the criteria on which its advice will be based are still up in the air.)

According to the report, the nine milestones remaining are: "Stress test system and demonstrate efficacy and accuracy; Assess accuracy of databases; Make modifications with respect to intrastate travel to accommodate states with unique air transportation needs; Establish effective oversight of system use and operation; Install operational safeguards to protect system from abuse; Install security measures to protect system from unauthorized access; Life-cycle costs and expenditure plans; Address all privacy concerns; Create redress process for passengers to correct erroneous information."

That's quite a shopping list, and it is hard to imagine a bureaucracy like DHS/TSA getting through it, even in the highly unlikely event that everything goes well. And if there are major problems, they will have to be identified and corrected, after the customary blame game has been enacted on Capitol Hill.

So the chance that this scheme will actually be implemented in August is very slim, especially when one considers the extraordinary capacities that it is expected to have.

An anti-terrorist machine

According to GAO, the required system capabilities are: "Comparison of data contained in the passenger's reservation (PNR) with information contained in government watch lists; Matching information in the PNR to CAPPS I rules to identify individuals who should be subject to additional security screening; Checking PNR data against commercial databases to assist in confirming the passenger's identity; Matching PNR data against lists of international fugitives and government 'wanted lists' to identify known criminals; Using algorithms developed through intelligence modeling to identify previously unknown terrorists; Maintaining a list of individuals, who have been previously cleared under credentialing programs, to minimize the volume of passengers that must be prescreened; Providing the capability to create a temporary watch list based on information extracted from current intelligence reports, such as blocks of stolen passports."

A computerized system that could reliably satisfy one of those requirements day-in, day-out, with passenger volumes as heavy as the USA's, would be mighty impressive. But here we have something out of Star Trek, in which taxpayers are investing billions, with so many potential points of failure that it will be a miracle if it doesn't increase the risk of hijackings.

It will be full of bad data that will repeatedly flag and inconvenience the wrong travellers. (The existing CAPPS system didn't stop the 9/11 hijackers, although it did catch US Senator Edward Kennedy and former singer Cat Stevens, for example.)

Worse, a system such as this is, by design, exceptionally easy for terrorists to reverse-engineer. By making a series of 'dry runs,' a terrorist crew can easily learn which members get flagged and which don't, insight of tremendous value for choosing the individuals most likely to succeed in an actual attack.

So it comes as no surprise that TSA should have fallen behind in developing a system intended to do the impossible. The only odd thing here is the fact that the law enforcement establishment, the public, and Congress foolishly persist in believing that "information technology" is the answer to real-world security problems. ®

Related stories

DHS comes clean on CAPPS, lets self off hook
ID theft is inescapable
Uncle Sam demands all air travel records
Airport snoop system thrown in $102m garbage can
Airport security failures justify snoop system

Sponsored: Today’s most dangerous security threats