Insurance Claims to be Tested by Lie Detector

Albanach writes "HBOS, one of the largest UK banks is to introduce random lie detector analysis of insurance claims according to this article from the Edinburgh Evening News. The three month trial will see calls from its 1.5 million policy holders randomly subjected to voice stress analysis. Those flagged up will then receive a set of questions designed to expose 'potential fraudsters'."

Comments

Yea right, I'm sure

[mjmalone]

they'll be chosen randomly, more like a 'you fit our demographics for a lieing bastard lie detector test.' Isn't this an illegal detainment/unjust search? What are the search/seizure laws in the UK anyways?

Re:Yea right, I'm sure

[oniony]

I really don't believe they would be allowed to do this without prior consent. Calls are typically prefixed with a "this call may be recorded for training purposes" but I doubt they would be able to do the same thing for detecting fraud.

More likely, one would have to consent in writing which they may offer a reduced premium to encourage people to sign up. (In reality, the money they save will possibly not be passed on: instead the 'reduced premium' could end up as being the usual price and those not signing up will pay a penalty).

Re:Yea right, I'm sure

[Anonymous Coward]

Unless I missed something, these are banks doing this not the government or police. And they aren't detaining you or performing any search or seizure. It is a phone system that looks for stress in your voice. I am sure it is pretty unreliable and not admissable in any court of law, though.

Or did you just mean - First Post?

Re:Yea right, I'm sure

[tdemark]

I've studied atmospheric phenomena for many years and I have yet to see bird contrails.

Perhaps the word you are looking for is this.

Although, I admit that bird contrails would be an interesting sight. =)

Re:Yea right, I'm sure

[vegetablespork]

I really don't believe they would be allowed to do this without prior consent.

That would be a condition of your policy, absent any consumer protection legislation to the contrary, which given the current U.S. administration is not bloody likely.

Re:Yea right, I'm sure

[rifter]

What I want to see is lie detector tests for lying insurance agents. :) Probably going to happen about when that law which dictates the flogging of spammers followed by their head being placed on a pike as a warning to others is finally passed...

Re:Yea right, I'm sure

[ozbon]

I'm more intrigued about the stress-levels part. They're testing it in the household insurance department, which means they'll be dealing with fun things like people who've just been burgled, or who've come home to find everything destroyed by fire/flood/whatever.

Surely under these circumstances, the voice-stress meters will be pegging all kinds of false positives and so on? I know I'd be all over the place, so any stress analysis is likely to be inherently flawed in such a situation...

Re:Yea right, I'm sure

[TheCarp]

However heres the thing... contrast that with the scenario of someone who wasn't burgled or who set the fire themselves and as such lost everything on purpose for the money.

I would bet dollars to donuts the stress would be different.

Or would it, I dunno, I supose I don't know much about what "Voice stress" really measures... but I can't imagine that a frauder and a real claim are going to be stressed in the same way.

-Steve

Re:Yea right, I'm sure

[operagost]

Considering that lie detector tests are still not accepted as evidence in court and likely never will be, insurance companies can go ahead and use polygraphs all they want. They still won't be able to use them to prove fraud or refuse benefits.

Silly

[grennis]

Wouldn't it be natural to have stress in your voice if something has happened in your life causing you to file an insurance claim?

Re:Silly

[ihummel]

And some people (like me) would find any such interrogation stressful.

Re:Silly

[ihummel]

Still, the very knowledge that I may be undergoing a lie-detector test that is imperfect enough to be inadmissible in court could cause significant stress.

Re:Silly

[mikerich]

Wouldn't it be natural to have stress in your voice if something has happened in your life causing you to file an insurance claim?

Not to mention stress induced by the 'Press 1 to speak to a human being who sounds like a machine, Press 2 to speak to a machine who sounds like a human being ... [BLIP]

You pressed 1. Press 1 to speak to a person in Edinburgh, Press 2 to speak to a person in Bangalore, Press 3 ... [BLIP]

You pressed 2. If you want to learn more about our low, low rates Press 1, If you want a cuddly toy as seen in our adverts Press 2, If you actually want to talk to someone Press 3 ... [BLIP]

You pressed 3. Are you sure you want to speak to someone? Press 1 ... [BLIP]

whirr clickity

Hello and welcome to the queue to join the queue to talk to one of our service representatives, you are number [pause] fifty-seven in the queue, estimated wait time is [long pause] - do you have any plans for October? While you are waiting, do you know about our other services?' rigamarole.

Best wishes,
Mike.

Re:Silly

[Foochar]

Thats why they do a baseline first. You start with questions that are pretty much guaranteed to be right. For example you confirm their name, their address, their phone number etc. This is used to establish a baseline for their voice stress levels. Its the same way that polygraphs work. You think if you are hooked up to all this equipment your pulse and resperation aren't going to be slightly elevated? Of course they are, but they establish a baseline first before asking the questions they are really interested in.

The other thing about this whole thing is that this isn't going to be the end of the process, its just one more thing to check. Kind of like if you file two claims too close together they take a closer look, even if both are legitimate. If you show more stress they are going to take a closer look.

Re:Silly

[madfgurtbn]

Thats why they do a baseline first. You start with questions that are pretty much guaranteed to be right.

Actually that' "baseline" is all bullshit.

The first questions to "calibrate" a polygraph are simply to mindfuck the person being tested. The user is told to tell the truth on a meaningless question, then told to lie on another meaningless question. The polygraph operator looks quizzically at the data then says something like "You're a terrible liar! It's off the charts!".

Lie detecting is fraud,whether by voice stress analysis or any other means. Read deeply at www.antipolygraph.org

It's about gaining confessions from gullible people. There is no real science involved. No lies are actually "detected".

Re:Silly

[Merk]

I'm lucky enough to have been polygraphed, but not for having done something important. I was actually a guinea pig for police officers being trained to become polygraph operators. I am not sure if polygraphs are completely worthless or not, but I can guarantee that the way they're used is not as a "lie detection device" but more as an interrogation technique.

When they polygraphed me, they used some cheap magicians tricks. They had me choose a random card, then told me to say no for each card when they asked "is this your card?". Using the polygraph they claimed to know what card I had chosen -- but the way they set things up it seemed more like magicians tricks, than it did polygraph operation. They also showed me a wavy line with a spike afterward, claiming that the spike was when they detected me lying... but when I tried to get some more details they avoided the subject.

When it came for the real test, it ran mostly like what you see in the movies. After the test was done they thanked me for cooperating and then started trying to usher me out. I asked if I could see the results and they refused. Later on, I was told that a polygraph operator never shows their results to anybody, not even their partners.

I'm not sure if the machines are completely worthless, or if they can do something, but it certainly isn't a lie detector, it's more a "reaction sensor" if anything. I wouldn't be surprised if the cops get more out of watching someone's face, eyes, and posture than they do out of the device. It's just that the device has such fame from TV and movies that guilty people think it will catch them, making them more nervous (and presumably innocent people feel it will vindicate them, making them more relaxed).

The only think I learned from the experience is that police interrogators are good at getting confessions. That doesn't necessarily mean they're good at getting only guilty people to confess, however. If you ever get accused of something and a cop wants to question you, whether you're completely guilty or completely innocent, insist on a lawyer.

Re:Silly

[Damned]

We covered this fairly well in a recent psychology and law seminar course I took.

The short version: The control question test is crap. The guilty knowledge test seems to be a good test.

Long version:

Polygraphs only measure autonomic nervous system response. A difference in their charts could be stress associated with lying, but it could equally be startle from a loud noise outside the interview room, uncomfortability with the particular question in general, annoyance that you've been asked the same question multiple times, or any number of other things.

As you stated, the baseline of the control question test (CQT) is worthless. It's all a trick to convince the testee(sp?) that the polygraph machine is an all powerful detector of false information. It is also very easily fooled by simple techniques. Polygraphers will say it has a 95% or better accuracy rate, but there is evidence that the CQT can have a one-third rate of false positives. The unreliability of the CQT is the reason why polygraph evidence is inadmissible in most courts.

The guilty knowledge test (GKT) is much more difficult to set up but more reliable in its results. The GKT works best in murder cases, and also in robbery cases as well. The police must keep certain aspects of the crime from the media, such as type of weapon used or where the body was found for example, and use those for the test.

The suspect is then attatched to a polygraph and asked questions such as "When you killed X, did you use a hunting knife, baseball bat, .45 pistol?" and responses are measured. Heightened responses, as well as the actual answer given, for the actual weapon used are interpreted as the suspect having knowledge that a particular weapon was used in the crime. Over a certain number of questions (I cannot recall the number) the probability that someone will have the highest reactions to the actual answers to the questions who did not have anything to do with the crime is very small, even if you were imagining how you would commit the crime.

From what I've seen, this test has a much higher reliability rate than the control question test though it is used about 0% of the time. It does, however, rely on the suspect not having been shown crime scene photos and the evidence to be used in the questioning not being leaked.

Why is it that I always think I've never made sense when I reach the end of anything I write?

Re:Silly

[guacamolefoo]

Classic ones:

1. Do you know why I stopped you?
2. We'll go easy on you if you just tell us what happened...
3. You don't ming if I look in your trunk, do you?

As far as the syllabus that you linked to, it looks like the classic "Did daddy touch you this way?" and then operator demonstrates with anatomically correct dolls. Kid goes "uh hunh."

The trick for adults is to shut the fuck up when a cop talks to them. Don't follow social rules about "filling silences" or "being nice". The cops are worse than telemarketers -- not only do they refuse to take "no" for an answer, they are frequently looking to arrest you for something. It's a lot worse than buying a subscription to a shitty magazine.

If questioned by a cop for anything you did not call them for (i.e. they are knocking and looking for you), here's the thing to do:

1. Say "I want to talk to my lawyer."
2. Shut your mouth.
3. Do not give them permission to do anything.

GF.

Great.

[ihummel]

Now they will be able to refuse you insurance payment based upon a method that doesn't hold up in court, at least not in the U.S. Does it in Europe?

Re:Great.

[colinleroy]

No, but as they plan to use this method only to find possible frauds, and investigate them using more standard methods, your point isn't valid.
Don't you think insurance companies have some lawyers paid to think about such issues before spending money on them ?

Re:Great.

[crizh]

CLAIMS are a huge problem in the insurance industry and all insurance companies do their utmost to avoid paying them.

Insurance companies employ people whose sole purpose in life is to find ways of weaseling out of perfectly legitimate claims.

That's disgusting.

There ought to be a consumer association that prepared figures of how much money each insurance company dodged paying to customers that thought they were covered (thanks to the persuasive sales staff) and ranked them.

That way, when you were thinking of buying a policy, you could check out the company and see how likely they are to actually pay out on a claim.

Insurance companies are wank*rs. Greedy, amoral, heartless wank*rs. Trust me.

Stress?

[BWJones]

So, when their customer who has just been involved in an auto accident calls and reports the accident to the insurance, their voice will not in the least be affected by stress?

It's a Manipulation Tactic

[Greenisus]

Lie detectors are not effective. This is just being used to scare people into thinking they can't lie.

Re:It's a Manipulation Tactic

[PurpleFloyd]

What happens if you get falsely accused? Lie detectors are at best random and can be conciously affected by an adept scammer (when you tell something you want to be recognized as the truth, relax, when you tell something you want to be detected as a lie, tense up a few muscles subtly; the toes are ideal as long as you wear shoes). Lie detectors are bogus science; while there may be physiological responses associated with dishonesty, they can be easily overwhelmed by other stresses (like going into a lie detector test and being grilled mercilessly). Also, it's very rare that two polygraph "experts" will read the results the same way. You might "fail" the test, while if your results had been interpreted by someone else, you might have "passed."

This is not just bad, this is awful. If you were falsely accused, you could land in jail and be out thousands of dollars in fines. Even if you miraculously avoided all that, you would still be left with a valid insurance claim that wouldn't be paid, despite the fact that you paid your premiums and did nothing wrong, other than fail a pseudoscientific test.

As for the supposed deterrent effect, that's a ridiculous analogy. You might as well suggest that we fine and jail people who "look suspicious" at random; you would get the same results. While you'd certainly catch criminals, you'd also punish a number of completely innocent people. Deterrent effect? No, there's a difference between deterrents and people living in fear of the law. The fact that polygraph tests are generally inadmissable in American and European courts should tell you something about what effects this would have.

Re:It's a Manipulation Tactic

[PurpleFloyd]

Let's take a hypothetical scenario:
You have a fairly new, expensive car with good comprehensive insurance. You leave the alarm system disabled by accident one day, and your car is stolen. You submit a claim to the insurance company and file a police report. A few days later, you are called into their office, to give a statement into a tape recorder. You fail the "voice stress analysis." Due to this, the insurance company starts to dig. They find that you left your alarm off, and think they can take you to court. You are taken to court and convicted of fraud, and punished accordingly. You are punished, but did nothing wrong.

What's the point I'm trying to make? It's simple: these essentially random tests will be used to determine who is suspected of a crime and thus investigated further, with a heavy bias towards criminal activity - investigators will tend to look for any evidence at all that might support the "criminal activity" theory, and doubt evidence that disproves that theory. It's a basic tenet of psychology that people tend to choose one theory and build up supporting evidence for it, while disregarding evidence that might disprove it.

Of the many cases detected by this "lie detector," there are almost certainly cases that have done nothing wrong, but have a large amount of circumstantial evidence against the person making the claim. While circumstantial evidence is technically inadmissable in court, expensive legal attack teams, like the ones held on retainer or employed by large companies like insurers and banks, can get away with almost anything and make it look reasonable. I doubt you could afford your own counterattack lawyers.

The end result is that it's possible for innocents to be punished. While I agree that insurance fraud is without a doubt a Bad Thing, and deterrents to insurance fraud are good, the chance of error here is simply too high.

Hopefully...

[rczyzewski]

It will decrease my insurance premiums. Probably not.

But what about...

[SUPAMODEL]

... the people who are actually stressed or whatever about making the claim itself?
I know lie detectors are supposed to be calibrated, but they aren't fool-proof and to hang decisions like this on them is just foolhardy.
Last time I had to make an insurance claim, it was against someone who thru their own negligent driving resulted in me having a serious enough motorbike accident to fuck my ankle, my bike & nearly write off their new, expensive enough, car. And I wasn't going fast, and did what I could to avoid it.
It was stressful enough having to deal with the claims people etc, tryin gto pick at everything anyway, so how is this going to help?

From what I have heard...

[zubernerd]

lie detector tests are about as good as flipping a coin. I wonder what the result would be if they just randomly chose ~50% of their claimants and investigated them...

This is bullshit

[rde]

Why not just get Uri Geller in to sense whether they're telling the truth? Time and again lie detectors have proved to be - at best - slightly better than waving chickens when it comes to accuracy. Yet all Liberty can say is "I'm not persuaded this works".

Re:This is bullshit

[gilesjuk]

"Hello, I'm ringing to claim on my insurance, I've had an accident in the kitchen and my spoons are all bent"

Train to beat it.

[eddy]

Antipolygraph.org

They're in Trouble

[Doesn't_Comment_Code]

Using voice stress analysis techniques to detect changes in speech patterns caused by stress, the machines will be able to make an initial assessment as to whether the caller may be lying.

A special series of questions has also been devised to try and catch out fraudsters.

And when was the accident?
Who was driving?
What's the capitol of Uzbekistan?
Pi to 15 digits?

I'm sorry sir, your claim has been denied.

This is news?

[henbane]

So insurance companies have found yet another way to bog down any legitimate claim you might have so they can keep the inflated premiums they took off you.

Any time they come up with BS like this they always claim it will lower premiums and give some inflated figure of how much fraudulent claims are costing them, but who is to say how many of those fraudulent claims are not just the companies finding a loophole to screw anybody who makes a claim.

Can't stand them, they have a business model where everybody has to give them money and they resent anybody with a legitimate claim to it.

Re:This is news?

[swb]

Insurance comapanies need to be divorced from the claims review process; they have a profit motive in rejecting claims. I wouldn't be surprised if there wasn't pressure on claims adjusters to increase their claim rejection rate or decrease payouts by small percentages when profits are threatened.

The best way to accomplish this would be to require a third party review claims, and if they find the claims to be valid, require the insurer to pay them. The third party claims reviewers would be held accountable for all insurer losses and costs associated with fraudulent claims, perhaps at 2x levels to increase their fraud scrutiny.

Hmmm...

[TWX]

Well, if they don't use the lie detector test as the only way of establishing truth or deception, this might not be completely terrible.

Remember, police don't generally use the polygraph to make a direct case against someone. They use it in conjunction with hard evidence and a narrowed list of suspects for a particular, established crime. As long as an insurance company is smart enough to not use the test to try to claim "you're lying! You weren't hit by the other driver" based on a nervous test taker who trips the system simply by knowing that (s)he is taking it, and they go on other evidence as well, like police reports and the like, things should be okay.

For other things, like theft, if someone is confident enough that they'll succeed by reporting something stolen, then trying to claim it on insurance, it's pretty likely that they'll now take steps to practice to lie to a polygraph convincingly. That would render things completely useless.

Either way, we'll have to see what the results of use are, and hope that they don't claim fraud upon people who are exhonerated later.

Lie detectors suck

[Goog500]

Lie detectors are not accurate. Somebody can be nervous because they're afraid of not being believed, and a well-rehearsed lie is easy to pass by a lie detector. Thats why lie detectors are not admissable evidence in court.

For the way technology has improved life press 1

[dirtmerchant]

"Thank you for calling HBOS Insurance. Your voice my be monitored to detect tiny fluctuations that may possibly indicate fraudulent statements. This technology is very controversial and invasive, but will allow us to prosecute one additional insurance fraud case each year. Rest assured, the money saved by fraud prevention will not be handed on to you the customer. Please hold for the next available underpaid outsourcer with no job security to copy all of your credit statistics into our closed-source database running the most up-to-date NT service pack from 1999."

Good.

[Chess_the_cat]

Homer: Well, let's see. There was the Picasso, my collection of fine wines, one or two Rolexes.
Insurance Agent: I'm sorry sir. This policy only covers real damage. Not made up damage.
Homer: D'oh!

There are a lot of people who try to scam the insurance companies and I end up paying for it through higher premiums. These are private companies and if you don't want to deal with one that uses lie detection don't. But I'll be the first in line to sign up with a company that does.

turnabout

[zornorph]

So will the insurance companies be willing to undergo lie detector tests themselves? If they are going to dish it out, they should be willing to take it.

Simpson's quote

[Lord_Slepnir]

Detective: Did you hold a grudge against Montgomery Burns?

Moe: No

Lie Detector: BZZZT!

Moe: All right, I did. But I didn't shoot him.

Lie Detector: Ding!

Detective: Checks out. All right, sir. You're free to go.

Moe: Good, because I have a hot date tonight.

Lie Detector: BZZZZZT!!!!

Moe: A date

Lie Detector: BZZZZZT!!!!

Moe: Dinner with friends.

Lie Detector: BZZZZZT!!!!

Moe: Dinner Alone.

Lie Detector: BZZZZZT!!!!

Moe: Watching TV

Lie Detector: BZZZZZT!!!!

Moe: All right! I'm going to sit at home and ogle the ladies in the Victoria's Secret catalog.

Lie Detector: BZZZZZT!!!!

Moe: Sears Catalogue.

Lie Detector: Ding!

Moe: Now will you unhook me already? I don't deserve this shabby treatment!

Lie Detector: BZZZZZT!!!!

Obligatory Simpsons quote

[DorkHead]

FBI agent Scully : This is just a simple lie-detector test. I'll ask some simple questions and you should answer with yes or no. Do you understand?
Homer : Yes.
[ The machine blows up ].

From the Skeptic's Dictionary:

[narcolepticjim]

From a page about the Truster Voice Stress Analyzer:

What is a voice stress analyzer, you might wonder? It is a machine that measures components of the human voice--frequency modulations--that are correlated with stress. No machine can detect stress directly, much less distinguish whether the stress is due to lying, guilt, stutter, fear, constipation, or some other emotion or physical condition. The frequency modulations, called "micro tremors" by those who measure them, must be interpreted by a human being. The machine doesn't do the analysis, the examiner does.

Some basic facts

[Anonymous Coward]

Here are some basic facts about insurance companies:-

1) They are out to make money
2) They make said money based purely on others suffering
3) They will try to weasel (no offence to weasels) out of *any* contract
4) Any money saved will *not* reduce premiums but increase bonuses and dividends
5) Insurance companies have never worried about legality. If they reject 50 claims (illegally) and only 5 have the time, energy and money to fight them they have made money on the other 45. All costs that the 5 have to pay, they cannot get back
6) Lie detectors are inammisable in UK courts - but that won't stop Insurance companies.


In the UK insurance companies work a "your a lieing defrauding piece of crap" policy. It's not even "guilty until proven innocent" policy. Most of the time Insurance companies believe _all_ people are trying to illegally claim.

Re:Some basic facts

[iabervon]

It's not an issue with calculating it, it's the fact that, while the odds are in your favor, the worst case is really bad. Would you rather buy something for $5, or have a 1 in 10,000 chance of paying $10,000 for it? Sure, the first is more expensive on average, but for the second, you have to be able to afford $10,000 in order to be safe. The insurance company is going to play 100,000 times, so they can take advantage of the fact that it will average out; you play only once, so it doesn't average out-- it's either good or really bad.

The other thing is that insurance companies get bulk discounts on a lot of the things that payments go towards. It costs them less to get your car repaired, because they have a lot of car repair business, than it would cost you for the same repair.

From nopolygraph.com

[Suhas]

Minnesota Polygraph Statute

181.75 Polygraph tests of employees or prospective employees
prohibited.
Subdivision 1. Prohibition, penalty. No employer or agent thereof
shall directly or indirectly solicit or require a polygraph, voice
stress analysis, or any test purporting to test the honesty of any
employee or prospective employee. No person shall sell to or
interpret for an employer or the employer's agent a test that the
person knows has been solicited or required by an employer or
agent to test the honesty of an employee or prospective employee.
An employer or agent or any person knowingly selling, administering,
or interpreting tests in violation of this section is guilty of
a misdemeanor. If an employee requests a polygraph test any
employer or agent administering the test shall inform the employee
that taking the test is voluntary.
Subd. 2. Investigations. The department of labor and industry
shall investigate suspected violations of this section. The department
may refer any evidence available concerning violations of
this section to the county attorney of the appropriate county,
who may with or without such reference, institute the appropriate
criminal proceedings under this section.
Subd. 3. Injunctive relief. In addition to the penalties provided by
law for violation of this section, specifically and generally, whether
or not injunctive relief is otherwise provided by law, the courts of
this state are vested with jurisdiction to prevent and restrain violations
of this section and to require the payment of civil penalties.
Whenever it shall appear to the satisfaction of the attorney general
that this section has been or is being violated, the attorney general
shall be entitled, on behalf of the state, to sue for and have injunctive
relief in any court of competent jurisdiction against any such
violation or threatened violation without abridging other penalties
provided by law.

dont need to.

[Unknown Poltroon]

Im a pathological lier. Polygraphs dont work on us anyway.

Re:dont need to.

[Col. Klink (retired)]

As George Costanza would say, "It's not a lie if you believe it."

Lie detectors don't always work

[Perdurabo26]

I've been submitted to a lie detector one time. Basically my house was broken into, and to clear my name, the police wanted to give me a lie detector test (i don't know off hand what kind it was) but i basically failed the test. The problem is that I failed the preliminary test too. I failed questions that were specifically designed to be correct.

Do you live in the state of michigan?

Do you live in the United States?

Are you 17 years old?

If you can't pass questions that are geared to be absolutely correct, than why do they still consider you failing the actual test as you lieing? I'm afraid something like this would happen in this situation.

Just giving my $0.02 worth.

Everyone Lies

[LittleGuy]

"Everyone lies, Michael. The innocent lie because they don't want to be blamed for something they did not do. The guilty lie because they don't have any other choice."
----- Sinclair, "Babylon 5", 'And the Sky Full of Stars'

Innocent until tested guilty

[eaolson]

I love this bit:

And Mr Hemingway said there will be measures in place to make sure only fraudsters are trapped, rather than those who naturally find making such phone calls difficult.

Apparently the system isn't capable of false positives. "You can't be innocent, the machine says you're guilty. And since only guilty people are caught by the machine, you can't be innocent. QED."

Also note that the article is talking about voice-analysis stress testing (over the phone, surely that couldn't ever be inaccurate), not polygraphs. Polygraphs are a crock as well, of course, but this isn't them.

Post-facto

[Johnny Mnemonic]

Funny they don't require this before they take your money. Maybe they should be subject to my lie detector when I subscribe with them: are they really going to be there 24 hours a day with a hold time of under 10 minutes? Will they settle all of my claims, or just the claims that they feel are reasonable? If a natural disaster occurs, and several $Billions are filed simultaneously, will I still be covered, or will they simply go bankrupt? Will they really save me 15% or more?

Really, brillant strategy. Take money, and then decline service later. Maybe computer techs should be in the same business: I'll take your money now, but when you need service I'll just blame it on you and continue to post to /.

Too bad it's voodoo science

[Hays]

On multiple occasions I failed polygraph tests that kept me from getting an internship. It's pretty annoying to have someone telling you you're lying. You're really quite powerless to do anything but deny it. Then they'll kindly show you the door.

These things have no place. They are not useful for job screening. They are not useful for investigative purposes. They are not reliable enough for any application. Congress was right to refuse to be polygraphed while under investigation- I would certainly refuse any future polygraph. They shouldn't be hypocrital, though. They should strike down polygraph use entirely.

Trusting polygraphs is a threat to our national security. Not only because double agents and such can easily pass them while lying (any well trained person can), but because so many qualified applicants are replaced with less qualified applicants who can satisfy the voodoo magic of a polygraph machine. Personally, I would like the very best working for the CIA, NSA, etc.

From the article:

[lungofish]

He added that it could also lead to lower premiums.

BZZZZZZZT! LIE DETECTED.

Re:why is this a problem?

[geekoid]

1) innocent people trigger lie detectors.

2) It becomes a 'prove your innocent' case if someone thinks you are lying.

3)people come to trust machines. So they will take a failable machine over a person.

4)the agent has to look at evicence and facts to determin wether you are lying.

5)IT put the burden of proof onto the victim.

Re:Insurance is a SHAM!

[Jonsey]

Actually, insurance really isn't. Note to pundits: I do work at an insurance company, albeit as a technology-related intern.

This year my company had it's highest amount of money made in a long time off of premiums. After cat damages (mass catastrophies are tracked differently due to reinsurance.... it's a long idea, and I'm not sure I could explain it well)... anyway, after all damages and claims paid out, our company made 96% return on premium.

That is to say, for each dollar collected in Premiums for policies, we paid out .96 cents.

This was astonishing for the finacial analyist! Most of the time, these numbers are around 114 or 108. That means on the property/casualaty side of the insurance agencies, the company loses money on all premiums collected. (Not as a result, but it's a good ratio to be able to judge the industry).

Anyway, Insurance companies make most of their money off of Investments in Bond/Stock markets. Insurance companies really are brokerages of a sort. Say your company has a 401k-ish plan. That money gets paid under a group header to an insurance company. That company, to make money for itself and you, invests in bonds (at my company) and stocks (at some others... we don't need huge returns with large sums of money, especially for huge risks). With the money the company has invested, it makes money.

Insurance is not a sham, it operates under the assumptions that finacial analysits can make more money than a random algorithm that buys stock/bonds/futures/money-market shares.

Note, I am a 2nd year intern here, I know insurance from presentations, and a mother who works in the industry. This information is not guarenteed to be true, but I'm pretty sure this is how it works. Take a look at quarterly earnings reports from Insurance companies... look at the combined ratios.

Insurance companies, at least my insurance company, are not out there to screw policy holders. We rather like ours. We don't like fraud, fraud is.... fraudulent. I didn't have a high opinion of insurance either until I found out how it worked.

I'm fine with that...

[FJ]

... as long as I can do the reverse. I want to make sure that when my insurance rep says "your covered" he doesn't mean "your covered as long as you never make a claim."

I also want to get a truthful answer to the question "Will I be dropped after my first claim?"

Nothing to Fear?

[CPIMatt]

"Honest policyholders will have nothing to fear and combating fraud will make things better for them anyway by helping to keep premium costs down."

Yeah, right. Honest policyholders do have something to fear; the fear that when this flim-flam pseudoscience piece of crap system randomly flags them as a liar!

Actually, the insurance company will probably use this like the police do. The insurance company will use the voice analysis technology against people who they already think are trying to defraud the company. They know it doesn't work worth a damn, but use people's ignorance of technology against them to get them to confess to faking a claim. The police do the same thing, they tell a suspect that they have failed a voice analysis test and use that as extra leverage to try to get that person to confess. It only works if the person is gullible enough to believe that the technology works.

-Matt

Im a semi-pro insurance fraudster ...

[Anonymous Coward]

so I'm posting anonymous coward for obvious reasons.
I think that this will catch low level fraud such as exagerated and false claims on things like household work and travel insurance. How well will depend on how good the equipment is and how frequently it is used. Also, publicising its use will have a great detterent effect which is why they are doing so.
However higher up the food chain this is just not going to replace effective investigation. In my scams I always ensure that I have employed some patsy like an office manager who believes what he saying when he talks to the insurance investigator - I do this because most of those guys are really good and pick up on body language which is a very frequent way they pick up on something phoney.
As the article says; it is only the start it will not by itself prove anything - it only provides a starting point. Those of us on the exciting side of the fraud industry know the key to a successful scam is thorough organisation and meticulous planning - most of my schemes would probably survive a 'red light' because of that, but I intend to continue avoiding that risk.

A useful toy but not the panacea thank God.

Your objections amaze me

[daBass]

Lie detectors are not evidence, they only lead to further investigations, which is fair enough.

From a crowd (/.) that not only advocates free speech but also the freedom to listen to anything that is being transmitted, yet you opose to a 3rd party listening in on your own transmissions?

Of course someone whose house has just burned down or car totaled will be stressed, but the evidence in these cases is so clear, that a police report can be trusted, something which can't be said about foreign police reports of many people who have the camera they don't like anymore "stolen" (as in donated to a nephew) on holiday. Claiming that, usualy days or weeks after the fact shouldn't put you under much stress, if it does and you can't come up with a good story to further questions, hesitating on too many details you hadn't thought about when concieving your fraud...

I am sceptical about the system but don't see this to be that much of an invasion of privacy.

It would deter real people

[burgburgburg]

There are a number of people who would be deterred from making legitimate claims because of fear of being falsely accused of scamming, general anxiety, concern over non-relevant liability issues. Of course, this would benefit the bottom line of the insurance company who would claim to have "stopped fraud".

And of course, real scammers will easily get by this. And since like most "security" measures, it make them watchers complacent: "Nobody is getting past our lie detectors."

Escaping the conventional polygraph.

[MirthScout]

I doubt these voice stress "lie detectors" are any better than the conventional polygraph tests. I participated in a study in college and found the conventional polygraph easy to beat.

Two years later while still in college I was working at a chain store in the local mall. Corporate headquarters "randomly" selected the store for polygraph testing of all employees. You can't refuse without being fired since you almost certainly agreed to polygraph tests as a condition of employment when you were hired.

When my turn came, I went in and sat down. The process was explained and I was handed some forms to sign. They basically said I wouldn't hold them responsible for any harm to me, my reputation, etc. from the test or their use of its results. It also said they could give the info to anyone they felt needed to know. I refused to sign the forms. I then held out my arm and said they have my permission to hook up the machine and proceed with the test.

That confused the hell out of them!

"I can't administer the test until you sign the form", he said. I told them that is not my problem. I aggree to the test. It is your decision to administer it or not.

They decided not to. They gave me a "security interview" instead. All the same questions but no polygraph machine. The man administering the test said that the security interview is actually more acurate because people are more relaxed and reveal more information.

Re:Complete Text of Article

[rifter]

"Honest policyholders will have nothing to fear and combating fraud will make things better for them anyway by helping to keep premium costs down."

Of course they leave out:

1) He was lying.

2) Since he is a practiced glib liar there was no hint of stress in his voice at all.

First off, I have to laugh when I read an article where lie detectors are described as "scientific" "sophisticated" and "accurate." They are frequently described as such, but it is clearly not the case. Even proponents, when pressed, always end up admitting that the "lie detector" is supposed to test stress levels; in other words, proponents of lie detectors usually lie in order to push them as a panacea.

Secondly, the biggest, oldest lie any insurance company can tell is that their rates are going to go down if you allow them to implement something. Insurance companies have historically tried to push legislation, promising practically every time that such legislation / policy change / newfound power will result in lower rates (mandatory automobile coverage comes to mind) but it never does. I go so far as to say I would think anyone would be hard pressed to come up with a single form of insurance in which rates have *ever* gone down, in fact.

Thirdly, the insurance company says that lie detector tests have been successful in reducing fraud. They do not qualify this at all, but I would think being able to point to a number and say "We were able to deny X million dollars worth of claims on the basis of lie detector tests alone!" would be considered a success, especially considering that the avoidance of paying claims, at any cost, any way they can is a goal to all insurance companies second only to raking in your cash.

The point they seem to be missing...

[The Angry Mick]

...is that most people generally have very good reasons for sounding distressed during a call an insurance company.

How is someone supposed to calmly explain they just lost their entire family to a car crash, saw their child die in a terrorist attack, or just permanently lost the use of their arm to the wood chipper? How are they supposed to do this while navigating the vast innefficient bureaucracy insurers have erected to keep callers to a minimum? Just getting through the bloody voice mail tree is often more than enough to send most folks into a rage, which'll probably light these lie detectors up like Times Square on New Year's Eve.

I get the feeling this is just another attempt for insurance companies to try and justify claim denials. Cheap and cruelly insensitive.

Typo

[t_allardyce]

Ok i think there was a typo in the story, they must have ment that they would use voice stress analysis on random calls and if it found that the claimer was too calm it would flag further investigation. If your house has just burnt down your probably not going to sound normal, if your planning a fraud you're more likely to have practiced a convincing conversation.

Insurance Claims to be Tested by Lie Detector...

[cbenesch]

Insurance Claims to be Tested by Lie Detector... Oh I had the faint hope that the policies they offer would be put under scrutiny. Hope springs eternal, Christian

What's next

[Nept]

a Voight-Kampf test?

Ask the NSA about it

[AndroidCat]

The NSA used to use heavy-duty polygraphs and experienced operators (possibly they still do), and still let some people pass which shouldn't have. (Possibly flunked some people which should have passed, but those people could hardly file a grevance.)

"Lie-detectors" are voodoo. Any informed court should tear a case based on those results to shreds. (Two weasel-words in there: informed and should.) I wonder how the insurance companies will hire trained and certified operators? Check for recent certs from the Cthurch of $cientology with E-meters?

My advice?
(a) refuse any such idiocy.
(b) if pressed, curl your toes on any tough question.