Hacker News Reader

Resampling DNA before conviction would fix the specific issue of the digital record being subject to tampering, I kind of assume this is done already.

The real nightmare is that we might be right around the corner from DNA being writable to the extent that you could make mock DNA to plant at a crime-scene/evidence-locker, and there might be a couple awkward decades where the courts believe the FBI when they claim to have found DNA matches (just like that time they systematically pretended they could identify hairs, and convicted people of murder based on hairs that were literally from dogs in some cases). We'll catch on eventually if they try it, but it would suck for those of us convicted on falsified evidence. I wouldn't be so paranoid about that possibility if they hadn't lied about the dog hairs for decades.

> right around the corner from DNA being writable

Why bother when it's trivial to copy existing DNA? The evildoer just needs to get a tiny sample from someone who'd made a decent frame.

Sure, then you've got the problem of how to create a DNA-bearing artifact that'll fool forensics -- a spilled puddle of pure chromosomes would be weird -- but that'd be true in either case.

I've read a couple books where the opposite was done. Intentionally dump huge amounts of DNA into the area to make any tracing pointless. One amusing and gross way was vacuuming bus seats for the material.

Keyboards are a great source of material. They accumulate no end of crap over time.

I wonder how often public library keyboards are cleaned, or the ones in university computer labs. When I was in college pulling all nighters in the computer lab I regularly saw the lab being cleaned, but I can't remember ever seeing the keyboards getting dusted off, let alone wiped off with disinfectants.

Go to two or three libraries in different corners of town, get what you want from the keyboards, and you should have quite the collection of DNA.

With the ease of travel I'd suggest go to libraries in different states or countries. Libraries close to major airports so that many people are likely to travel between them. (ie don't pick a city in the middle of nowhere just large enough to have a private airport)

Fun activity: turn the keyboard upside down over a sheet of newsprint or butcher paper, and tap the front corner gently on a hard surface.

The word for this type of accumulation of mixed biological material, crumbs, dandruff, eyelashes, and such is "kroonch".

Reminds me of a scene in The Town - they were going to rob a bank, and they collected whole bags of hair from a barbershop floor, and spread it during the robbery.

Wouldn't that be quite useful for police?

They may be able to pinpoint a location - the barber where all the DNA matches go to have a haircut, and a time - soon after all those with majority DNA in the pile of hair went to the Barber's. Then they can analyse who was in the area around the time and find a person involved.

My thoughts exactly. The only true way is to appear to have never been there forensically, leave no sort of evidence. Security through obscurity only works in some instances.

There was a murder in Nebraska where crime scene was liberally cleaned with bleach to obliterate DNA evidence. Still no conviction despite a lot of circumstantial evidence. This was partially compounded by older DNA technology that destroyed the sample while processing it.

> This was partially compounded by older DNA technology that destroyed the sample while processing it.

Can you point to current DNA tech which doesn't destroy the sample while processing it?

For PCR you only need a super-tiny bit, then you can amplify it as much as you want. I guess you're still technically destroying that tiny bit, but it hardly makes a difference when you only need a few molecules.

That wouldn't actually do anything for DNA, cut hair doesn't actually contain any DNA at all the only way to get DNA from hair is if the follicle is still attached which wouldn't be true generally in a barber shop.

Not true anymore.

https://www.nytimes.com/2019/09/16/science/hair-dna-murder.h...

The Unabomber retrieved pubic hair from a public restroom to plant as a false clue in one of his bombs[0].

0: https://www.chicagotribune.com/news/ct-xpm-1998-05-05-980505...

Sometimes the best way to hide is in plain sight.

Indeed! We did Polymerase Chain Reaction experiments when I was in senior biology in high school almost 20 years ago.

$150 for 10 groups: Edvotek 330 Amplification of DNA by PCR, For 10 Lab Groups https://www.amazon.com/dp/B004W7ZAYI

Because if you make an oligo from someone's digital data it's harder to link you - you may never have been on the same continent together.

It's like snatching handbags vs online credit card fraud - the reach of criminals to victims got a lot bigger.

This is already here: https://www.forensicmag.com/news/2015/02/dna-evidence-can-be...

DNA has long lost its aura of invincibility.

A partial or even complete match is not enough to convict anyone.

And with technology like CRISPR, you kinda wonder how that will impact forensic science.

Not at all, probably.

From what I know about CRISPR, and forensic DNA testing, it should be extremely difficult or impossible to, for example, change your "genetic fingerprint" as it is called.

If I remember correctly, those methods primarily use repetitive patterns, which should be hard to modify with CRISPR. And you'd need to modify hundreds of them, and in billions of cells of your body (even if you try to only get white blood cells and their progenitors).

That sounds like something nation states would be able to do well before it's acknowledged.

Another nightmare, it's already possible to create convincing audio samples of someone else's voice that can be used for framing. I'm sure believable, fake video footage is around the corner too. Then what? You can create fake audio/video of anyone committing crimes. Or it can be used to fabricate an alibi for someone who is actually guilty. Not sure where we go after that. Maybe back to the time before audio/video technology was invented and rely solely on witness testimony.

Deepfaked videos are here already, I could see one passing as believable evidence (especially if it's shrunk down to a lower resolution, as in a high quality deepfake pretending to be a cellphone video).

We had this brief period of history where good records of actual events were cheap to capture and expensive/impossible to fake (governments have been airbrushing photos for most of this period, but videos seem to have been tougher). We're at the end of that now, but the public hasn't quite caught on. I think it'll take at least one high profile case getting overturned before society at large stops believing in videos. Then we're essentially back to trusting people at their word/digital-signature, and we'll probably have "trusted" security camera companies that install tamper-proof cameras at multiple angles that cryptographically sign videos while streaming them to remote servers. Before that, and probably after it as well, some of us will rot behind bars based on state-forged evidence.

People said the same thing when photo-editing software was new. How is this any different?

> rely solely on witness testimony.

Bad part about that - even today - is that we know for a fact that what people remember/recall isn't necessarily what actually occurred. Also, as time goes on, those recollections can become more unreliable.

We know that the mind confabulates, making things up to fill in gaps - both at the time something occurs, and after the fact to "fill in the story".

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4183265/

https://faculty.sites.uci.edu/starklab/false-memory-eyewitne...

Unfortunately, the legal system has been slow in acknowledging this - on top of so many other things found and/or known to be false or bias - two big ones are so-called "lie detector" systems, and "blood spatter analysis".

Fortunately, lie detector "evidence" is generally not admissible in the United States (at least I think for criminal cases?) - but blood spatter analysis is still considered to be a "real science", when it may be anything but:

https://features.propublica.org/blood-spatter/blood-spatter-...

I know that members of the military already have been entered into DNA databases, and there are reasons to do this with police officers and firefighters. Mostly to identify them postmortem.

I don't subscribe to DNA information being a secret. It tends to be "secret" in the sense that it is non-trivial to obtain. But you leave behind a trail of DNA everywhere. Ultimately, it cannot be held secret.

Therefor it is more important to regulate the use of such information, for example to prevent discrimination.

i suspect the military/leo databases are not the ones being searched in criminal cases

In fact when your DNA is taken for the military database it is explicitly stated that it will not be used for criminal cases. If anyone managed to search the database, that would certainly get thrown out by a court.

> In fact when your DNA is taken for the military database it is explicitly stated that it will not be used for criminal cases. If anyone managed to search the database, that would certainly get thrown out by a court.

I wouldn't be so sure. Evidence observed during the execution of a warrant against the incorrect property is still valid and admissable, for example. Not saying this kind of thing applies in this specific case but thought the idea worth mentioning.

Aren't there exceptions when there is a reasonable cause to suspect a member of the military?

Law enforcement in the US could build a pretty powerful genetic genealogy database just using their own DNA.

I don't think these are typically used (or could be used) to convict someone, I think it's mostly used to guide investigations.

This is correct, it's used to find suspects, not to convict anyone.

AKA fishing

If done right, spear fishing.

I'm a privacy guy, but the science really is there to use DNA to narrow a list of suspects. You can't say the same for much of forensic science, or for that matter mass surveillance technologies.

However, the list of suspects can still lack the actual criminal(s), but still contain an innocent with poor, but true alibi.

Here's an actual case of an innocent person who faced charges and a high probability of losing their freedom due to DNA transference:

https://www.wired.com/story/dna-transfer-framed-murder/

The take-away is that shed skin cells containing DNA can migrate and a DNA match from that type of evidence shouldn't be relied on as proof of anything.

To my view, the combination of DNA transference, advancing DNA collection / analysis techniques, the mere existence of DNA databases, and familial searches are absolutely terrifying.

Good counterpoint, definitely an issue. When the police find macroscopic quantities of bodily fluids, however, it would seem to be madness not to use the evidence.

Absolutely.

I do question if society is better off catching a few more criminals through these databases and familial matching vs the potential for innocents to be imprisoned based on nothing more than a DNA match against skin cells.

I'm also not too fond of the idea of my relatives' DNA being used to draw suspicion to another relative just because someone's DNA wound up in a database. To me it's just another one of those things (ALPRs, all the compilations of "it's just metadata") that erode our privacy and give the state a great deal of power for relatively little benefit to the people.

All crimes don't get solved, all criminals don't get caught, and in a society where everyone's a law-breaker because there are too many laws to know them all and so many of them are dumb, that's how things should be.

"Fishing" in police work is kinda the opposite:

You investigate someone, hoping to find some law they might have broken.

This is trying to find out who committed an already known serious crime. That's just "police work", in my view.

Except of course that possibly the only link between the suspects and the crime is the DNA, they may not have committed anything.

But once such a suspect is lined up the search is over and all that will happen is to look for confirmation rather than exculpation.

Any suspect can always be innocent.

There is nothing special about DNA evidence that makes it worse or better in that regard.

You dont have a warrant to find those peoples things.

Get a warrant before searching for each persons DNA, not search each persons DNA then get a warrant

Why wouldn't they be allowed to be used? Anything that guides an investigation is supposed to be admissable evidence or else you get into fruit of the poisonous tree territory, which is illegal. Not saying it wouldn't happen still, but they wouldn't be talking about using it in that case. It'd be like with stingrays where they publicly claim that they got a tip from a confidential informant and drop cases on anyone who gets close to the truth so the never actually get a judge to tell them they can't do it.

> Anything that guides an investigation is supposed to be admissable evidence or else you get into fruit of the poisonous tree territory, which is illegal.

Slight correction here.

The FPT doctrine makes it so most[0] evidence obtained illegally cannot be admitted. It doesn't work in reverse, it does not imply that all inadmissible evidence is illegal.

For example, the first half of many investigations is full of inadmissible hearsay, before you narrow in on what happened and find the people who actually saw something and can testify. There's nothing wrong with talking to people to figure out who knows things, and see how stories changed when they were told to different people, so you can more fully understand what happened.

How this applies to your main question --

It might be very difficult to use a DNA database company results in a US court because you might have to involve a ton of witnesses to vouch for every part of the process. Suppose the defense asked, "But how do we know the post office didn't mix up different samples that day!?" You would be forced to either (a) have the USPS provide testimony establishing a full chain of custody, or (b) just test the defendant now to verify the match. It will always be easier to do (b), so I'd guess[1] everyone just skips to that step.

[0] Even FPT has exceptions...

[1] Maybe there are other procedural ways to approach this that are more common now, I haven't been involved in cases like these or researched it specifically, my criminal procedure knowledge is getting really dated so I might be wrong here.

> > Anything that guides an investigation is supposed to be admissable evidence or else you get into fruit of the poisonous tree territory, which is illegal.

> Slight correction here.

> The FPT doctrine makes it so most[0] evidence obtained illegally cannot be admitted. It doesn't work in reverse, it does not imply that all inadmissible evidence is illegal.

> For example, the first half of many investigations is full of inadmissible hearsay, before you narrow in on what happened and find the people who actually saw something and can testify. There's nothing wrong with talking to people to figure out who knows things, and see how stories changed when they were told to different people, so you can more fully understand what happened.

> How this applies to your main question --

> It might be very difficult to use a DNA database company results in a US court because you might have to involve a ton of witnesses to vouch for every part of the process. Suppose the defense asked, "But how do we know the post office didn't mix up different samples that day!?" You would be forced to either (a) have the USPS provide testimony establishing a full chain of custody, or (b) just test the defendant now to verify the match. It will always be easier to do (b), so I'd guess[1] everyone just skips to that step.

I never thought of this but it would be surprising if the USPS _didn't_ maintain some sort of chain of custody of mail items, at least at the bag level or whatever unit they use.

I would be super surprised if they _did_ that seems like a massive amount of work for not much gain.

Maybe something like logistics (i.e., bag went out for delivery on truck FOO) but I doubt they track everyone that has had access to it.

USPS has been scanning all mail going through sorting facilities with cameras for years now. All mail is photographed, image goes into a database, image is OCR processed, etc. This is only in recent years, but still. Interesting stuff: https://www.newsweek.com/postal-service-photographs-every-pi...

> I never thought of this but it would be surprising if the USPS _didn't_ maintain some sort of chain of custody of mail items, at least at the bag level or whatever unit they use.

Probably so. And in civil contract law we almost presume the USPS works more or less like a black box and always functions correctly. I mean, more specifically here, you might not have to question the USPS but might have to question the lab techs, or some company representative, and open up debates about what safeguards they use to prevent samples from getting mixed up. Lots of openings you could avoid by just testing on site, so that will always be easier. You generally just do whatever you can to remove extraneous witnesses to simplify your argument.

The subtle trick with US criminal law is that it is still a "testimonial" system. You have to have someone vouch for every piece of evidence. If you take photos of the crime scene you technically have to ask someone questions like, "do these represent the scene of the crime?" ("How would you even know? What makes you an expert on what this place looks like?") Some of this stuff can get waived but in contentious cases you dot your is and cross your ts.

Physical evidence is a bit like a hacked on extension of an old oral only court system. You originally just asked a bunch of villagers what happened and hoped they were sufficiently scared of swearing oaths to not lie.

Unless they use parallel construction after the fact...

https://en.wikipedia.org/wiki/Parallel_construction

Which is the illegal part.

But also a hard to discover part and a hard to fight part.

That does not seem to be enough to stop it from happening.

The whole second half of my original comment is on how even if it was illegal, we still wouldn't be seeing them talking about it so publicly, comparing it to their use of stingrays which is pretty much veiled in secrecy.

"unbelievable" on Netflix is a good example of such leads- family tree DNA leads to two brothers

It seems like any competent defense attorney would insist on resampling, or alternatively insist this evidence thrown out duebto lack of chain of custody.

Signed, not a lawyer.

Insisting on resampling would be stupid, it will match twice that is not the problem. But if you insist on resampling then you are giving the evidence more weight than it can bear, much better to show that statistically it does not have much weight at all even if it matches (once, twice or more is not relevant).

The cold cases broken open are the ones that are left after all the other methods have already been exhausted. Sure, some of those might have remained unsolved if not for this method. But that does nothing to show the method by itself works better than all the other police work, for that to be concluded you'd have to compare with all the cases that were solved with ordinary police work and then to conclude that if all of them were first attempted to be solved with this method that the total outcome would be an improvement.

That seems unlikely. So for a very small fraction of extra cases solved a very large change in evidence gathering methods and dragnet style fishing for suspects does not seem to be worth it to me.

Stupid if your client is guilty and you trust the database, not so stupid if your client insists they are innocent and the database governance processes are suspect.

That test will run the same way twice unless it has been tampered with, you'd have to be pretty sure before taking a gamble on that. The chances of a false positive are >> the chances of database tampering, why attack the harder target?

If you have somebody in custody, you can double-check the DNA record with a new cheek swab.

The problem with fishing expeditions on DNA databases is that when you run a 'one in a million false positive' test against a database of 300,000,000 people, you're going to get 300 false positives.

More than one of those suckers just might not have an alibi.

I am curious about the actual false positive rate and how they compound. To use your hypothetical number, if the false positive rate is one in a million per test, if law enforcement tests your DNA sample against 700,000 samples from crime scenes, there is a 50% chance at least one of them will be a match due to false positive [1]. That is much scarier than what one in a million suggests.

[1] (1 - 1e-6) ^ 700,000 = 49.6 % chance of no false positive match

Or you now have 299,999,700 probable negatives, and 300 possible positives to review for motive, method, and opportunity before you start looking at/for alibis.

Unless you've got a cold case, that seems like more potential suspects than a general person may reasonably have.

Oh yeah, lots of American police want to work this hard; there's probably five or six of them at least. What most police will do instead is arrest the first name on the list. When she or he documents an alibi, they'll arrest the second name on the list. Eventually, some poor bastard who lives in the same state will have gone hiking alone the day of the crime, or stayed home and slept after a week of overtime, or whatever. Unless there is a statistician on the jury, that person is going to prison.

Statisticians probably get excluded in jury selection, along with anybody else whose profession indicates they have a clue.

https://www.quora.com/Are-engineers-usually-excused-from-jur...

That's like saying police shouldn't be allowed to use Google's "similar image" search because there could be false positives among the similar images.

These genetic databases are tools used for investigation to identify suspects. Once the suspect is identified, the investigators still have to build a case against them, including obtaining their DNA directly for testing against the crime scene DNA. Since all the cases I've read about the police got the suspect's DNA samples from the garbage, it doesn't seem like the genetic research itself is even enough to get a warrant for the suspect's DNA.

The hypothesis of this example is that the unfortunate suspect's DNA "matches" the crime scene DNA. Of course it's going to "match" when they test it again. The point is that when the original match is against a huge population, you're going to find a bunch of matches, most or all of whom are innocent. Picking one out of that set and then confirming that one is really in that set doesn't prove anything.

I'm not an expert, but I would guess it would be possible to have an exact match on the genetic fingerprinting, but then do a full genome sequencing to confirm that the two samples are from different people.

The funny thing is that although I have often heard about the police finding an exact DNA match with someone who was provably not at the crime scene, I've never heard about anyone doing a different DNA test afterwards to investigate what happened: did a fragment of skin get carried in the wind from one city to another, did samples get mixed up in a laboratory, or was the match a false one, akin to a hash collision?

These DNA tests do not use full genome sequencing, they use SNPs.

https://en.wikipedia.org/wiki/SNP_genotyping

I agree with your conclusion in general, but you add in some very basic secondary checks and you won't get that many false positives... do they live anywhere near the crime? Do they have any motive or opportunity to commit the crime? Were they in the area at the time?

So now you are fishing with a net with smaller holes. You're going to catch suspects that look bad because you've selected for that but that still may not have anything to do with the crime you are investigating.

God forbid any of them are minorities and the victim a white person, you're 95% on the road to a conviction.

But if you're testing 300Mi most of those are probably not even in you jurisdiction or have no ties to the case