The officers are busy. Last year nearly 800 cyber tips filtered through the National Center for Missing & Exploited Children. Officers are carrying out search warrants on a weekly basis and some electronic files have been spread so widely that officers know their hash value by heart.
It’s difficult to say whether sexual exploitation and the spread of sex abuse materials has actually increased dramatically in recent years.
This whole article feels like a cute dog story with barely any journalism about something that is very threatening to the Internet. Was it was submitted to trigger debate about propaganda rather than on the merits of the reporting?
I imagine there is a lot of useful statistical analysis to be done on the list of filenames, their metadata, and various content hashes which could absolutely answer some questions about the patterms of CSAM over the past decade.
Network effects are definitely interesting to law enforcement. There was an interesting podcast about one of the sites that got busted which was being run out of Korea: https://darknetdiaries.com/episode/131/
I went to RSA in San Francisco. At the time, I was staying with friends in San Mateo, but I got a hotel for a night in SF. Rates were outrageous, and we were a small startup. The cheapest hotel I could find was $400/ night for a bedroom with a shared bath down the hall.
The next night, back at my friend’s house, I noticed the telltale 3 bite pattern on my shoulder. Breakfast lunch and dinner. It itched like CrAZy. It was bedbugs. There was no way to know if I’d only gotten bit at the hotel or if I’d brought bugs to my friend’s home with me.
I learned you can hire a sweet yellow lab like Mojo to sniff out bedbugs. The dog wandered the whole house. My heart dropped when he alerted on the sofa where I’d been remotely working.
It cost about $3500 to treat the house. They rolled in fans that heat the whole house above 105°F for an hour. Sure enough, they found some bedbug eggs. It all worked out, damn what a hassle.
Note: the dog was from a different company than the treatment folks. They do that to avoid conflict of interest.
You need to get 113F/45C to kill bed bugs and eggs within 90 minutes. 105F actually sounds too low. But its harder to kill eggs, which you need to get up to 118F. They are resillient and evil creatures.
Not only that, police dogs are notoriously not great at actually sniffing out drugs instead of just alerting on meat products or on people their handler is biased against.
> You're no innocent. Pollack. An honest citizen is content with an ordinary data set like yours there." She pointed across the living room at the forty-by-fifty-centimeter data set. It was the great-grandchild of the old CRT's. With color and twenty-line-per-millimeter resolution, it was the standard of government offices and the more conservative industries. There was a visible layer of dust on Pollack's model. The femcop moved quickly across the living room and poked into the drawers under the picture window. Her maroon business suit revealed a thin and angular figure. "An honest citizen would settle for a standard processor and a few thousand megabytes of fast storage." With some superior intuition she pulled open the center drawer — right under the marijuana plants — to reveal at least five hundred cubic centimeters of optical memory, neatly racked and threaded through to the next drawer which held correspondingly powerful CPUs. Even so, it was nothing compared to the gear he had buried under the house.
Buried would make it harder for the data storage detector dogs to find.
FWIW, that forty-by-fifty-centimeter display has a 64 cm diagonal or 25.2 inches, and the twenty-line-per-millimeter is 508 pixels/inch or slightly above a good Retina display. That sounds pretty decent.
On the other hand, a few gigabytes of fast storage sounds ridiculously small.
>Mojo can quickly sniff out devices like hard drives, flash drives or SD cards — often leading officers straight to the evidence they need to make an arrest.
Do nand flash chips and magnetic platters have a distinct smell? lol :)
I'm not saying its impossible but really really outlandish
Yes! I saw a presentation a couple of years back from police dog trainers. Different dogs would specialise in different things - drugs, guns, and amazingly - hard disks! They train them specifically to do so. What blew my mind was the reward was always just a tennis ball. But the dog was soooo happy to get the ball for a short amount of play when it did well.
Drug-sniffing dogs are a myth[1]. They're used as a pretext for illegal searches, not to detect drugs.
I suspect this is similar. "Detecting off-gassing" is vague enough to sound like it makes sense, but there's a variety of materials in electronic devices and many of the ones that off-gas (like plastic) are found in millions of other products.
Your citation pretty clearly demonstrates that drug-sniffing dogs DO work. And anybody who has ever had a dog sniff something out would find this to be an obvious fact.
The omitted figure is how large the sample size is. If the dog sniffs 100,000, and identifies 10 as having drug odor, but only 2 of the 10 are currently in the possession of drugs, that doesn't really indicate that the dog "has a failure rate of 75%" because it doesn't take into account any of the "negative identifications." We don't know how many of those other identifications were false negatives, but certainly I wouldn't expect 25% of the population to be carrying illicit drugs at any given time, and especially not in scenarios where there may be a checkpoint with a dog involved.
The dogs may be misused, but that doesn't make their ability a myth, and neither does this article claim that a dog can't sniff out the trace scents of a drug. It carefully skirts around this assertion, because dogs absolutely can sniff out drugs. They're just not infallible, and may also be used as pretext.
But in the case of drugs there would be a clear motivation fuelling the alleged myth: to perform illegal searches based on the fake prior of "illegal drugs" having been sniffed by the dogs. That does not work for electronics.
Carrying or distributing certain drugs is illegal, and thus a sniffer dog ostensibly detecting them can be grounds for reasonable suspicion and a search. Being suspected of carrying electronics, on the other hand, is not grounds for a search. That is, of course, unless these dogs are claimed to be trained to sniff out only electronics storing child pornography... ¯\_(ツ)_/¯
Well, there's a chance that people had used drugs or smoked before, and the smell might still be present. The article acknowledges this, so it's not entirely a myth. While dogs can make mistakes, it's not definitively incorrect.
That's not how the dogs are used. They are told to signal, it's not a mistake. It's a tool of intimidation and a performative gesture towards probable cause. Officers without dogs will just claim "smells like weed" or "your eyes are bloodshot" or a million other things.
When actually searching for drugs, they prefer to destroy the vehicle outright, kick in any box speakers, tear off door panels and fabric, cut open seats, pull out spare tires, etc. They never ever use a dog to find the location of drugs in the car, only for the excuse that drugs might be in the car.
Of course I've never done a search, I'm not a cop... I've been searched many times though, and what I described, I've seen and experienced. Maybe you've done too many trainings and don't know what happens in real life?
> Whether or not the dogs were "incorrect" was a definitional question for Dr Malins.
The dog was accurate, however the dogs definition of accurate (there is some minimal smell) vs. the enforment officers (i need some smell so i have a ground for search) vs. the people (i walked through a cloud of smell) might be very different, which makes the measure "accurate" - well - very inaccurate.
> Do nand flash chips and magnetic platters have a distinct smell?
Not chips. I'm electronics engineer with large experience. You will not believe, how many extremely non-nature chemicals used in electronics. Number is really huge.
And these chemicals really have strong smell, when touch with soldering iron I feel it even when powerful ventilation working, and I've tried to feel smell even when finished device and washed it with alcohol - something left anyway for months.
For about storage, I'm not sure, but have hypothesis - electronics likes volume and some fabs extremely specialized - for example, 80% of storage in whole world made in just THREE fabs (also exist fabs specialized on CPUs, on radio, on sound amplifiers and other analog circuits), and they might have their own unique list of used chemicals, and dog could detect this list and could feel difference from other lists.
Nintendo uses bittrex in the plastic for their Switch cartridges to stop kids from swallowing them. Potentially this is more widespread in the SD card industry? (I don’t have any around me to test-lick unfortunately.)
I'm also concerned. The only reason I can guess at would be to prevent data from being seized when going through a security crossing, such as when going from one country to another or into/out of a secured corporate facility.
Maybe we can be charitable and assume that the parent commenter did undercover work as a journalist/mule in an authoritarian regime, eg. bringing photos that North Korea didn't want to be public out of the country? They're not likely to let you leave with an encrypted uSD even if you claim you forgot the password. But the vast majority of reasons to employ such deception are more nefarious - either illegal corporate espionage or military espionage across an airgapped/monitored network, or distribution of CSAM.
Or maybe he just likes the taste of microSD cards?
I manage a fleet of over 1500 raspberry pis and sometimes put microsd cards in the same pocket I put my supplements and just scoop the pocket full and put it in my mouth and drink some water and swallow
After that happened once I did it again deliberately in front of some friends because it was funny
The linked-to article in turn links to https://indianapublicmedia.org/news/indiana-man-training-k9-... dated January 27, 2017 describing how ""I had to pay a chemist to actually find the actual odor," Jordan says. "So it's an odor that's within the actual SD cards, the thumb drives, cell phones.""
This is further detailed at https://www.techrepublic.com/article/electronics-sniffing-do... where "[Dr. Jack] Hubball examined hard drives, thumb drives, SD drives–every type of electronic storage device available. The common denominator? A circuit board."
> He began testing various circuit board components, and about six months later, identified a compound called triphenylphosphine oxide (TPPO)–which covers the circuit boards in all storage devices from large hard drives down to microSD cards to keep them from overheating.
> Another compound, hydroxycyclohexyl phenyl ketone (HPK), was extracted from removable media, such as CDs, DVDs, Blu-Rays, and even floppy disks.
> Dogs trained to sniff electronic equipment are the least known specialization of canine training, which has been implemented after the 9/11 terrorist attacks. International Police Forces begun to train search dogs for the retrieval of electronic equipment such as USB sticks, micro SIM cards, mobile phones, DVDs, CD-ROMs, external hard drives, and memory cards. Specialized dog units in this sector are called Electronic Storage Detection Dogs (ESDD) [24]. The ESDD dog is trained to sniff out the chemical component, triphenylphosphine oxide (TPPO), common to most electronic gadgets and storage devices, as a thermal insulator [25].
> Analyses found several volatile compounds common to SIM and SD cards, as well as USB drives, including 2-propenenitrile, styrene, isophorone, hydroxycyclohexyl phenyl ketone, and 2-furanmethanol, tetrahydro. Results indicated that mass storage devices do have a characteristic odor profile making detection with minimal false alerts feasible for trained canines.
Note: this is a feasibility study only - it does not test if dogs could do it. Nor does it mention the TTPO mentioned in the Italy paper, which is also not mentioned in [25] -- which had nothing to do the the topic!; see
Another lead is https://www.taylorfrancis.com/chapters/edit/10.1201/97810032... "This chapter introduces the project within the National Police of The Netherlands and focuses on the training method used for “Digital Storage Device Dogs.” Canines have previously shown to be an effective tool in locating obscured digital storage devices in the United States and United Kingdom." I don't have access to the content.
A news blurb about the UK use is at https://ieeexplore.ieee.org/document/8401991 but does not include what substance is detected. It does say the dogs used in the project came out of the Connecticut State Police work.
I would like to know of dogs which have been trained for this task which did not come out of that one program, and/or learn more about the specific tests done under more rigorous circumstances than the ones described in the links I gave.
Still, it does seem feasible that it works as described.
Anyone who's worked in electronics manufacturing knows that printed circuit boards and soldered assemblies have a distinctive smell. The epoxy laminate boards get washed, but still have an odor, and we wash the flux residue off after soldering, but there are still odors that linger when acres of PCBs go across your manufacturing floor.
I am a little surprised that a dog's nose is sensitive and rapid enough to find the circuit board in a microSD card. The number of atoms of this residue that off-gas and waft through the air to the dog's nose must be tiny!
Sounds like one just needs to coat absolutely everything in their places, neighborhoods, businesses, etc. in TPPO and the like to change the noise floor.
> The workload has flooded,” Raymond told VTDigger. “And I think a lot of it is the availability of it online, obviously, but the other part is just, I think platforms are getting much better at controlling their own environment and finding and rooting that stuff out
A lot of that stuff will be much harder to find out if end-to-end encryption stays legal. Is it still worth it? Very possibly, but it's definitely a tradeoff and we shouldn't pretend otherwise. We're trading off real children being actually hurt against unspecified hypothetical harms in the future. Those harms may be much more significant, or they may not, but we shouldn't pretend like it's all roses.
This technology protects everyone equally, and more importantly will exist regardless of whether or not you want it to. Any regulatory attempt to kill FDE will only result in the use of more deniable systems by criminals and the mass exploitation of everyone else who relies on the physical security of their computing devices. The tech isn’t magic. We must simply live with the consequences of the reality we exist in, and adapt accordingly in the hunt for those who prey on children.
The fact that the storage media that this dog can find contain anything but Veracrypt volumes proves that (most) criminals have terrible opsec. Most people who own CSAM are ordinary people, teachers, plumbers, truck drivers etc, not hackers, programmers and IT technicians.
Child abuse is just like everything else, low-friction solutions are going to win over high-friction ones. If you have to get an Android phone and sideload a shady app to get E2E encryption, many people won't even bother. In the world where E2E is banned, half of the shady E2E apps would be police sting-ops in the first place, and the other half could be seized and turned against their users with a strategically placed update. Just like with money laundering, you don't even have to prove that a crime occurred, simply getting the list of all IP addresses of people using E2E would go a long way in putting those people behind bars.
Again, I'm not saying that this is necessarily a good thing, but it's undeniable that many children would be saved here.
Requiring all residential buildings have at least a single camera in every room that’s active 24/7 and is streamed directly to NSA(or whoever) would also save a lot of victims of crimes.
You could have AI detecting instances of domestic abuse and immediately dispatching a police drone to handle it. Technologically this is almost already viable..
You can justify anything with that line of thinking. To anyone paying attention I think it's obvious that AI image generators are going to end this problem. No need to hurt anyone and people can look at all the depraved things they want in the comfort of their homes.
Just because it's not possible to study something directly doesn't mean that we can't study it. We could compare statistics in countries with different laws. Or we could look at studies of how pornography consumption influences non-deviant sexual behavior.
Even the “end to end encryption” option for iCloud (off by default, mind you) keeps the raw hashes of the unencrypted file content non-e2ee so Apple/USG can see who has which unique files and to where they spread.
It’s difficult to say whether sexual exploitation and the spread of sex abuse materials has actually increased dramatically in recent years.
This whole article feels like a cute dog story with barely any journalism about something that is very threatening to the Internet. Was it was submitted to trigger debate about propaganda rather than on the merits of the reporting?
I imagine there is a lot of useful statistical analysis to be done on the list of filenames, their metadata, and various content hashes which could absolutely answer some questions about the patterms of CSAM over the past decade.