Airport screening doesn't work... don't tell anyone!

| Comments (3) | Security: Airport
Colorado's 9News reports on the TSA's Red Team tests of the screeners at DIA. They're not doing the most impressive job:
The Transportation Security Administration (TSA) screeners failed most of the covert tests because of human error, sources told 9NEWS. Alarms went off on the machines, but sources said screeners violated TSA standard operating procedures and did not hand-search suspicious luggage, wand, or pat down the undercover agents.

The Red Team uses very expensive chemical simulates in the test devices that look, smell and taste like real explosives, except they do not explode. To the CTX bomb detection machines at DIA, they are real explosives, according to a former Red Team leader.

Sources told 9NEWS the Red Team was able to sneak about 90 percent of simulated weapons past checkpoint screeners in Denver. In the baggage area, screeners caught one explosive device that was packed in a suitcase. However later, screeners in the baggage area missed a book bomb, according to sources.

Of course the TSA says this test is unrepresentative, but that this kind of result should be kept secret:

Morris says other agents, not with the Red Team, test and train screeners every day at the nation's 450 airports and says screeners pass most of those tests. In those kinds of tests, he said Denver has done well in the past.

However, tests done by the Department of Homeland Security's Office of Inspector General and the U.S. Government Accountability Office in 2006 found widespread failures. According to the GAO, screeners at 15 airports missed 90 percent of the explosives and guns agents tried to sneak past checkpoints.

...

Most test results, including results from the Red Team, are secret, classified as SSI or sensitive security information. Morris says they do not make them public because they could point out holes in the system.

So, there are two types of information here: the first is that the security screening has an incredibly high false negative rate. The second is the specific things you could do to get past security. It's certainly true that knowing specific ways to exploit security would be useful to terrorists, but it's not clear that the Red Team did anything particularly surprising or sophisticated here. In one of the tests, the agent appears to merely have outbluffed the screeners.

Now, one could argue that the mere fact that screening is so inaccurate is in and of itself useful to the terrorists, since this makes airports a more attractive target. But then this is hardly secret information. First, GAO tests showing very similar results have already been published. Second, all you have to do is know how the screening technology works (and that's no secret) and watch how screening is performed to know that it's not going to work that well. On the other hand, it's perfectly clear why TSA would wish to keep such embarassing information secret.

3 Comments

The way things are in the US these days, probably less than 10% of the people carrying guns or explosives are actual terrorists. (The rest are just your average US citizens.) So even if they can only catch 10% of the people trying to smuggle something on to a plane, the TSA figures that's still enought that they can catch all of the terrorists, with some room to spare. That makes sense, right?!?!?

Surely there's some middle ground here. The Red Team could reveal that the screeners performed terribly--earned an F-, or whatever--without revealing anything specific (like a 90% failure rate). I'd think that that would accomplish the goal of generating political and bureaucratic pressure to improve matters, without revealing anything of value to terrorists.

Surprisingly, I tend to agree with Dan. Revealing the failure rate and the sample size seems to be valuable information. If I were a terrorist leader, they would allow me to plan more accurately. A letter grade seems like a fine compromise. Not that it's a huge obfuscation, but it could conceivably help a bit.

Leave a comment