US Supreme Court Rules on Real False Positives

Last year when US Senator Ted Kennedy was detained for being on a no-fly list, I discussed his plight in relation to intrusion detection system "false positives." If an IDS is operating correctly, every alert it sees is the result of an action it was programmed to take. In other words, when a functioning IDS sees "cmd.exe", it reports seeing "cmd.exe".

It doesn't matter if the appearance of "cmd.exe" on the wire is not part of an actual intrusion; a rule to alert on "cmd.exe" does not cause "false positives" if the IDS reports seeing "cmd.exe". A real false positive involves the IDS reporting "cmd.exe" when no such content passed on the wire. Therefore, there are no such things as false positives. Blame the signature writer or IDS developer, not the IDS.

Let's move from the realm of IDS false positives to the land of canine false positives. Yesterday the US Supreme Court issued its opinion in the case of ILLINOIS, PETITIONER v. ROY I. CABALLES. This is a case where false positives involve a dog's ability to sniff for illegal drugs. Justice Ginsberg's dissenting opinion summarizes the facts of the case:

"Illinois State Police Trooper Daniel Gillette stopped Roy Caballes for driving 71 miles per hour in a zone with a posted speed limit of 65 miles per hour. Trooper Craig Graham of the Drug Interdiction Team heard on the radio that Trooper Gillette was making a traffic stop. Although Gillette requested no aid, Graham decided to come to the scene to conduct a dog sniff.

Gillette informed Caballes that he was speeding and asked for the usual documents–driver’s license, car registration, and proof of insurance. Caballes promptly provided the requested documents but refused to consent to a search of his vehicle. After calling his dispatcher to check on the validity of Caballes’ license and for outstanding warrants, Gillette returned to his vehicle to write Caballes a warning ticket. Interrupted by a radio call on an unrelated matter, Gillette was still writing the ticket when Trooper Graham arrived with his drug-detection dog.

Graham walked the dog around the car, the dog alerted at Caballes’ trunk, and, after opening the trunk, the troopers found marijuana."

Justice Stevens' majority opinion held that "the dog sniff was performed on the exterior of respondent's car while he was lawfully seized for a traffic violation. Any intrusion on respondent's privacy expectations does not rise to the level of a constitutionally cognizable infringement... A dog sniff conducted during a concededly lawful traffic stop that reveals no information other than the location of a substance that no individual has any right to possess does not violate the Fourth Amendment." In other words, it's ok for police to use dogs to inspect cars for drugs during traffic violation stops (or at other times), even if there is no suspicion of drugs involved.

I do not agree with this opinion, for several reasons. The first reason involves false positives, and was correctly diagnosed in Justice David Souter's dissenting opinion:

"I would hold that using the dog for the purposes of determining the presence of marijuana in the car’s trunk was a search unauthorized as an incident of the speeding stop and unjustified on any other ground...

The infallible dog, however, is a creature of legal fiction... [T]heir supposed infallibility is belied by judicial opinions describing well-trained animals sniffing and alerting with less than perfect accuracy, whether owing to errors by their handlers, the limitations of the dogs themselves, or even the pervasive contamination of currency by cocaine...

In practical terms, the evidence is clear that the dog that alerts hundreds of times will be wrong dozens of times.

Once the dog’s fallibility is recognized, however... the sniff alert does not necessarily signal hidden contraband, and opening the container or enclosed space whose emanations the dog has sensed will not necessarily reveal contraband or any other evidence of crime."

Justice Ginsberg expresses the second reason for my disagreement. Returning to her dissent, we see that beyond a Fourth Amendment violation, there are other problems with allowing canine searches prone to false positives:

"A drug-detection dog is an intimidating animal... Injecting such an animal into a routine traffic stop changes the character of the encounter between the police and the motorist. The stop becomes broader, more adversarial, and (in at least some cases) longer. Caballes –- who, as far as Troopers Gillette and Graham knew, was guilty solely of driving six miles per hour over the speed limit -– was exposed to the embarrassment and intimidation of being investigated, on a public thoroughfare, for drugs...

Under today’s decision, every traffic stop could become an occasion to call in the dogs, to the distress and embarrassment of the law-abiding population...

Today’s decision... clears the way for suspicionless, dog-accompanied drug sweeps of parked cars along sidewalks and in parking lots... Nor would motorists have constitutional grounds for complaint should police with dogs, stationed at long traffic lights, circle cars waiting for the red signal to turn green."

My third and final reason for disagreeing with the Court's opinion is based on Justice Stevens' majority opinion. He writes for the Court:

"We have held that any interest in possessing contraband cannot be deemed 'legitimate,' and thus, governmental conduct that only reveals the possession of contraband 'compromises no legitimate privacy interest.'"

Now, what if the definition of contraband is extended beyond illegal drugs? How about music or movies in digital form, or pirated software? Is the Court opening the door to knock down privacy rights, since means to discover contraband do not infringe Fourth Amendment rights? The Court continues:

"The legitimate expectation that information about perfectly lawful activity will remain private is categorically distinguishable from respondent’s hopes or expectations concerning the nondetection of contraband in the trunk of his car."

The Court also brushes aside the false positive concerns:

"Although respondent argues that the error rates, particularly the existence of false positives, call into question the premise that drug-detection dogs alert only to contraband, the record contains no evidence or findings that support his argument."

I find this ruling very disturbing. I expect to see canine units used in increasing numbers in the coming months, where false positives will continue to plague innocent people. For example, yesterday National Public Radio reported that a man carrying cash to close on his house purchase was arrested when a dog alerted to supposed traces of illegal drugs on the money. Apparently traces of drugs on US currency is not an urban legend!


Anonymous said…
I doubt that there will be any increase in the frequency in the use of drug sniffing dogs at traffic stops. This is something that has gone on for a long time, its only now that someone has brought it to the Supreme Court to rule on it. The ruling only states that use of drug sniffing, a resource that law enforcement has only a finite amount of, in this instance does not violate the Fourth Amendment. It only adds that State and Local laws do not trample the rights granted to citizens. Drug sniffing dogs will not be included in every squad car, nor will there be K9 units dispatched to the scene of every traffic stop for one very simple reason, resources. There are only a limited number of K9 units, and that’s assuming that department X even has K9 units to begin with. And yes, there are false positives with dogs, but there are false positives with polygraphs as well, and IDS systems. This ruling indicates that the use of the drug dog in this case was a passive indicator that something malicious has taken place, not an intrusive one, and not a violation of the rights of citizen. Remember, indicators and warnings are a signal that further investigating is necessary.
Anonymous said…
I'm not familiar with any aspect of the case you are reffering to but the facts as you describe them suggest a chain of events more likely than a random K9 search. Being pulled over for traveling six miles per hour over the posted limit is not unheard of but is not that common either. That pretext for the stop in conjunction with the officer's request to search the vehicle implies he was suspicious in the first place but did not have enough to perform a search without permission. I really doubt the K9 unit just happened to be in the area at the time. I suspect the officer found a creative way to confirm his or her suspicions within the letter of the law. As with all forms of security, even basic law enforcement, your activities are sometimes directed and influenced by what can only be described as a "hunch". Over my fifteen some odd years as a security practitioner, I've learned to trust mine because they're usually right.

It's a huge jump from getting creative to test a hunch and walking drug dogs around cars at stop lights. I can't image any K9 unit thinking that would be a good idea!
Note that Justice Ginsburg's statement said "Although Gillette requested no aid, Graham decided to come to the scene to conduct a dog sniff." This sounds like a bored cop to me, not one acting out of suspicion or probable cause.

Thanks for your comments!
Anonymous said…
Actually no, it probably was not a bored cop. This is a common practice in law enforcement, so it is not uncommon for an officer to provide assistance at a traffic stop. On the subject of false positives, a K9, although not infallible, does give rise to "reasonable doubt". An alert from the dog does give the indication that something may be amiss. Just as an IDS system gives an alert to a possible incident that warrants further investigation by an analyst, a drug-sniffing dog does the same thing. The details of this were not discussed in your post, but the officer had to get an order from a judge to search the car, or get consent from the driver. If he had not, this case would not have gotten to the Supreme Court and would have been thrown out at a lower court. When going before the judge (probably with a phone call), the officer only needs to give his reasoning as the alert from the dog to be granted the search order. The courts recognize drug-sniffing dogs as a valid indication of “reasonable doubt”. In the past, there have been attempts to "suppress the evidence" which have been dismissed by judges, so this is nothing new. Also the details of the dog and his handler were not given. He may be a drug dog that has been through the courts already and been recognized, nor do we know the history of the dog and his success rate in detecting contraband. He may have a 100% conviction rate in cases where he has detected narcotics. These things play a factor in a judge’s decision to issue a search warrant based on the indication given by a dog, and these details were not given in the article. Given that all of these issues have been addressed at lower courts numerous times with good reasoning, and also by more educated individuals in the area of law than you or I, I’m not surprised the Supreme Court would not overturn these rulings.

As far as the officer’s presence, remember he is allowed in any "legal" place. You have to remember, he is a "peace officer" and he is backing up another "peace officer" at a traffic stop.

Besides all of that, in this case this was not a false positive. This is a true positive and as far as I can tell, the ruling does not address the issue of false positives since that is out of the scope of the Supreme Court to rule on. The Supreme Courts function is only to rule on the constitutionality of laws. And besides, I only read from this the ruling pertains to the use of drug dogs and if it violates the Fourth Amendment, which if I am not mistaken, has been ruled on in the past.
"Radical" Russ said…
Hmm, a car pulled over for going a mere 6 MPH over the limit? I wonder what the car looked like. Wanna bet it wasn't a Lexus, Jaguar, or BMW? Oh, a driver named "Caballes"? I wonder how dark his skin tone is.

The fact that the case involved a "true positive" drug alert is irrelevant. This decision sets precedent. This gives police free reign to bring on the dogs whenever they have a "hunch" about a driver's drug use. Drivers of VW buses, cars with Grateful Dead stickers, and any used car leaving a Dave Matthews concert, you better get used to dogs sniffing around your car.
Anonymous said…
Caballes is a Spanish name, and he is Filipino. And he was driving a 1998 Grand Marquis. Not a very suspiscious looking vehicle, considering the arrest was made in 98... Read the details of the case before you speculate and pull the "race card".
Regarding this comment:

"The details of this were not discussed in your post, but the officer had to get an order from a judge to search the car, or get consent from the driver. If he had not, this case would not have gotten to the Supreme Court and would have been thrown out at a lower court. When going before the judge (probably with a phone call), the officer only needs to give his reasoning as the alert from the dog to be granted the search order."

The Supreme Court opinions state that Caballas did not give consent. There is nothing in any of the Supreme Court opinions regarding getting approval from a judge. The Illinois Supreme Court decision says nothing about getting approval from a judge either. It appears the cops denied Caballas' refusal to give consent based on the dog's alerting action.

Justice Souter's dissent provides details on the unreliability of drug dogs. Law enforcement officials have also shared similar concerns with me about drug dog reliability. That is a major problem for me; unreliable indicators are being used as the basis for investigating potentially law-abiding people. This is similar to labelling any bank transaction over $10,000 as suspicious per the Bank Secrecy Act (.pdf).
Anonymous said…
I am doing a Mock trial for class and the case involves a drug sniffing dog. Basically, the arresting officer came upon a civilian, with their car on the side of the room with emergency lights flashing, and cones to warn oncoming traffic "Chris Smith" finishing up changing a car tire do to a flat. It was a humid hot night and "Chris Smith" was sweating, the Officer asked if Chris needed help. Chris refused, stating" No, I am almost done." So the officer then asked" May I search your car" Chris then stated " No, I know my rights" with that the officer called for backup, and a drug sniffing dog came onto the scene, and started sniffing Chris's car, where he argued" This is an illegal search". But the Officer said" It is not illegal unless the dog are officer's touch his car".

I am the defense attorney in the case, and this seems like some Bull. The officer had no reason to call for backup, nor suspect drugs, nor was this traffic stop, or traffic violation. My client, Chris Smith, was acting legally, even more reasonably then most who handle flat tires.

So my question is, what right does that officer have to treat this like a traffic stop? The officer was acting purely on prejudges because of my client’s youthfulness rebellious.
Anonymous said…
I doubt that the justices can lay claim to a higher degree of accuracy and integrity than can a trained police canine.
In a case where a vehicle is parked or stopped on a public street, it is acceptable to walk a canine around the vehicle. This falls under the plain view doctrine (plain smell) that allows an officer that is lawfully present and observes unlawful activity, to act. The dog has the same authority and a dog sniff is NOT a search.
Anonymous said…
If a dog sniff results in an indication from the dog, you do not need judicial authority to search. This is (and should be) probable cause.
Anonymous said…
This comment has been removed by a blog administrator.
Justice David Souter said, "In practical terms, the evidence is clear that the dog that alerts hundreds of times will be wrong dozens of times."

Not true.

It could be MUCH less accurate. A dog that alerts hundreds of times may be wrong hundreds of times.

The dog-handler officer in Paster Stevens trial, shown at, testified that he keeps no record of when his dog alerts to drugs or smuggled passengers and a search finds no drugs or smuggled passengers. He further clarified that he has no idea how often his dog alerts in situations without actual drugs or smuggled passengers, though he did say his dog never "false alerts" because his dog always responds to the SCENT of illegal drugs or smuggled passengers. (He did not say how he confirmed that such scents were always there on alerts.)
Re the mock trial involving "Chris Smith":
If a reasonable person in "Chris Smith's" situation would have felt that he was not free to leave, then the officer was detaining "Chris Smith" for Fourth Amendment purposes. Detaining someone longer than needed for the original stop (or allowed by consent or by probable cause for a vehicle) is unconstitutional. There's a U.S. Supreme Court case on that, though I forget the cite.

Popular posts from this blog

Zeek in Action Videos

New Book! The Best of TaoSecurity Blog, Volume 4

MITRE ATT&CK Tactics Are Not Tactics