ACLU sues TSA to make it explain junk science "behavioral detection" program

Originally published at:

1 Like

That whole “holding government accountable” thing hasn’t really been working out, has it?


It’s pretty hard to do when the government cries “but, but, but National security” anytime citizens try to hold it accountable.

[edit] wrong pronoun.


[quote=“doctorow, post:1, topic:54205”]TSA refuses to explain how it spent $1B on a discredited “behavioral detection” program that led airport authoritarians to believe that when they racially profiled fliers[/quote]Probably by making a small number of clever people very rich.


As an NLP Practitioner I use micro expression analysis regularly with clients and as a sociological hobby (people watching).

The problem with the TSA isn’t the behavioral detection, it’s with the intelligence of the employees.

1 Like

Grifters gotta grift. I don’t really blame Paul Ekman for taking the TSA’s money, because Hey! Free money!

The supporting evidence might turn out to be “We watched ‘Lie to Me’ and it totally worked there!”


The behavioral detection is rather a lot of the problem. It is trying to extrapolate relevant security details from “expressions” which are ambiguous. The expressions themselves are not problem behavior, they indirectly imply other problem behavior.

The stickier legal problem is that I am legally culpable only for my actions, not my emotional states. Because my emotional states do not directly lead to unlawful actions, and I have negligible control over them. The micro expressions are considered “valid” precisely because they are involuntary, which eliminates them from having any legal weight. It’s a lot easier to argue that a person can’t fly because of a criminal record, or something in their luggage - compared to having too much tension in one of their jaw muscles.

I’d argue that people don’t need to be policed. But if it does happen, they need to be policed based upon what they do, rather than what they feel.


Why was this bullshit “behavioral detection” ever funded? The TSA can legally grope my crotch to check for explosives. What actionable information could possibly be obtained from also exploring my micro-expressions?


The usual budgetary story – spend it or lose it.

1 Like

Or even worse, oh you got by with $$$ less this year so you won’t need that $$$ in your budget next year.

This type of tool can only be useful in conjunction with other skill sets, and should only be an indicator, not a sign of guilt.

1 Like

This is all loaded. Useful to whom? Indicative of what?

I tend to be, in my daily life, hyper-vigilent. There are numerous times I have had people deliberately attempt to take my life. So I always expect - deliberately - that anybody, anywhere, may present to me some sort of conflict, possibly armed, and possibly fatal. I do this as a survival strategy, because I’ve needed it. I am always “poised”. Having armed thugs look me over doesn’t do anything to put my guard down. I might need to kill these people, or die trying. And I am certainly not interested in having this position hypocritically critiqued by law enforcement or soldiers when they are trained to do the same thing. They have their risk assessment, I have mine.

So, there are laws (for better or worse) which stipulate how what I do, or not do, effect my ability and expectations to travel. I am not interested in “indicators” of risk when I am not doing anything unlawful. And when I am harassed, the risk increases.

1 Like

Thanks for clarifying.

I am a psychotherapist and so I was commenting from a diagnostic position.

to me, TSA employees aren’t qualified to diagnose mood or intent and as such giving them such tools is pointless and a waste of money.

1 Like

I agree. But I think that it’s a deeper situational and legal issue that even if they were qualified to diagnose mood or intent, there is not sufficient legal framework to allow for them acting upon this. And if somebody wasn’t prepared to act upon these diagnostics, such systems would not be employed there.


At some level, the “problem” is the incredible rarity of terrorists. It simply doesn’t matter whether TSA agents can be trained in techniques that are likely to spot terrorists. Because fewer than 1 in 1,000,000 flyers is a terrorist, the number of false positives will mean that more than 1,000 non-terrorists will be wrongly identified for every terrorist. And a 1/1,000 chance hardly qualifies as the “reasonable suspicion” required to detain and question somebody.


To establish the rarity of terrorists, people would need to agree on a definition of the term. Something the US stubbornly refuses to do with any precision. Leaving the definition wide open as current standards are just gives the US “license” to harass whoever they like. The US even meets their own definitions of terrorism.

A good start might be to use a more less sensational term. How about “insurgent”? “combatant”? Something for which a functional definition can be agreed…

1 Like

It’s probably more than just that…

and also…

and also…

NLP is woo. When used as a justification by TSA to covertly apply racial profiling, it’s toxic woo.


I’m not clear what you are saying there. Analysis of people’s expressions is not a form of “programming” or hypnosis. In this sort of domain, it would be considered kinesics, a communications discipline which has been used for decades. But traditionally this has been done in interviews/interrogations for clues on how to proceed in the investigation of a crime which has already happened. Using it to profile masses of people early in hopes of detecting criminal intent is far more dubious, because it’s too open to interpretation. And because the hypothetical crime has never actually occurred.

See the third wiki citation. “NLP Practitioner” can mean absolutely anything, and the empirical evidence to support the theories is sorely lacking.


Ok, so?

I haven’t looked into it myself, so I could only others evidence on faith.