U.S. Relies on Admittedly Bogus Lie Detector in War on Terror

AntiPolygraph.org ^ | Dec. 2, 2004 | self

[George Maschke]

The website of the National Institute of Truth Verification, which markets the Computerized Voice Stress Analyzer (CVSA) -- a purported voice-based lie detector -- confirms that the U.S. Government is using this device for intelligence purposes in the war on terror. And yet the manufacturer has admitted in court that the device "is not capable of lie detection." See, Federal Use of CVSA Confirmed.

[Prime Choice]

Gee...why aren't the antipolygraph.org guys running the War on Terror? They seem to fancy themselves as being oh-so-much smarter than the dumb hicks at the Pentagon...

[endthematrix]

Ever heard of http://brainwavescience.com/ ?

[George Maschke]

Yes.

[endthematrix]

Your lying.

[Wiz]

Oh no! Now the government is going to sue them for an unreliable product.

What about the Truster, Handy Truster and other voice lie detector? Are they unreliable too?

Truster
Handy Truster

[George Maschke]

No, I'm not. :-)

[George Maschke]

All voice stress lie detectors are unreliable. None of them are supported by any peer-reviewed research. It's only the purveyors (and operators) of these devices who claim they can detect deception.

[endthematrix]

Heh, "You're LYING!" (get it...Lie detector?)

[Wiz]

Typo Lie Detector

[Lie][][][][X][][][][True]

Uncertain :)

[George Maschke]

I'm not sure I get your humor. But perhaps you're getting at a very salient point about lie detectors. Although they have no scientific basis, and cannot actually detect lies, they can be useful as an interrogatory prop. The interrogator questions the suspect/prisoner/intelligence source and then informs him/her that this highly precise instrument indicates with virtual 100% certainty that they are not telling the full truth. Gullible people who believe that the lie detector has read their mind sometimes make admissions/confessions.

But what is scary is that it seems that government officials (especially in DoD) actually believe in their lie detectors. To actually rely on the results of a voice stress analysis "test" (or a polygraph, for that matter) to determine whether or not to trust someone is incredibly foolish.

[this_ol_patriot]

They hooked it up to Bubba and it said he was lying so they thought it worked.

Think about the above statement.

[George Maschke]

Prime Choice,

I'm not claiming to be smarter than anyone. But it is a fact that the U.S. Government is relying on a lie detector in the War on Terror that the manufacturer has admitted is "incapable of lie detection." Doesn't that bother you?

[asgardshill]

But it is a fact that the U.S. Government is relying on a lie detector in the War on Terror

The only "fact" presented here is that this National Institute of Truth Verification is claiming on their web site that the government is "relying on" this so-called lie detector. Do you believe everything you read on the Internet? And unless the National Institute of Truth Verification has a mole working at JCS or on the theater commander's general staff, they're hardly in any position to have any knowledge about what the government is or is not "relying on", now are they?

Further, do you really believe that the government would just let an Al Zarkawi or even Bin Laden himself walk out the door with an "oh, so sorry" and a new suit of clothes if this spiffy box with the blinking lights indicated that they were not lying during an interrogation?

[George Maschke]

I certainly do not believe everything I read on the Internet. Yet I see no reason to doubt the National Institute of Truth Verification's statement that "Since August, 2003, [they] have trained nearly 100 CVSA Examiners for DoD elements, and provided more than 60 CVSA systems which are being employed by various DoD elements in the Global War On Terrorism."

This is an established company whose customer base, before August 2003, had consisted almost exclusively of state and local law enforcement agencies. They have little reason to publish false claims that could fairly easily be disproven, and potentially much to lose thereby.

I'm not suggesting that Zarqawi or Bin Laden, if captured, might be released based on Computerized Voice Stress Analysis results. That's a straw man argument you're making.

The problem to which I'm calling attention arises when U.S. intelligence personnel allow the results of an admittedly unreliable test to influence their decisions regarding the reliability of information received from prisoners, informants, prospective employees, etc.

[asgardshill]

I find plenty of reasons to doubt the affirmative statement you made that the government is "allow(ing) the results of an admittedly unreliable test to influence their decisions".

First doubt: The assumption that these boxes are the sole source of determining the veracity of interrogation results is ludicrous at best. If you honestly believe that these blinky-light boxes are the sole intelligence-gathering asset used by the government in this capacity, I have a bridge in Brooklyn I know you would be interested in buying.

Second doubt: The NITV simply is in no position to confirm the claims they are making on their web site. NITV is a private company - they are not the Secretary of Defense, the JCS, or a theater commander. In short, they have no way of knowing the government's actual policy on this or how much confidence the government actually has in the results emitted by their products.

NITV's used-car salesman puffery that their products "... are being employed by various DoD elements in the Global War On Terrorism" could also be made by any commercial supplier from Hanes Underwear to the Heinz Catsup Company.

The problem to which I'm calling attention arises when U.S. intelligence personnel allow the results of an admittedly unreliable test to influence their decisions regarding the reliability of information received from prisoners, informants, prospective employees, etc.

Third doubt: You've provided no real proof other than a few statements on a web site that the government is relying on these results as authoritative.

This is an established company whose customer base, before August 2003, had consisted almost exclusively of state and local law enforcement agencies. They have little reason to publish false claims that could fairly easily be disproven, and potentially much to lose thereby.

If NITV really wants to cut their own financial throats by making these claims, they're certainly free to do so.

I'm not suggesting that Zarqawi or Bin Laden, if captured, might be released based on Computerized Voice Stress Analysis results. That's a straw man argument you're making.

It was part of the whole gestalt that requires that I believe that a private company is somehow in the position of knowing military and intelligence policy in the War on Terror.

[George Maschke]

It is indeed an inference on my part that the U.S. Government is allowing the results of CVSA tests to influence its decisionmaking. But I think it is a reasonable inference. Based on NITV's statement (which I, unlike you, accept at face value) that they have trained nearly 100 DoD personnel and delivered more than 60 CVSA systems, I infer that these are being used as more than just an interrogational prop to bluff interrogees in an attempt to elicit admissions/confessions. (If such were the case, there would be no need to go to all the expense of buying NITV's systems and training; one could simply use any computer with a microphone attached and tell those being interrogated that it's a lie detector.)

I think you reach too far in supposing that a company that provides both lie detectors and training to the federal government couldn't know how its devices are being used. There is no a priori reason for such to be the case.

Note that apart from NITV's statement, the notion that voice stress analysis is being used in intelligence operations is also supported by the testimony of a released Guantanamo detainee.

[Buckhead]

It is not a Perry Mason moment of triumph for critics of this technology for the manufacturer to admit that a voice stress analyzer does not detect lies. It detects stress. The presence of stress is a factor which the examiner considers in determining whether the subject is lying based on the tendency of most, but not all, people to experience stress when they lie. The stress can have other sources, which is why voice stress analyzers cannot be said to detect lies.

[ko_kyi]

Any normal person who is asked:

Is your name Joe

Do you own a car

Did you commit crime X

Is going to know that the third question is the important one and if their livelihood is at stake is going to experience stress regardless of the truth of the answer.

[George Maschke]

The NITV touts CVSA as a tool for determining whether or not a person has spoken the truth. Their admission that it cannot detect lies directly undermines these claims.

Moreover, it has not been shown through any peer-reviewed study that CVSA reliably detects stress, let alone deception...