57 comments

[ 3.2 ms ] story [ 104 ms ] thread
The creepiest bit is the GSK story: Where even if I've been careful to not use any genetic testing services, not granted anyone permission to access or use my medical data, etc., I'm still screwed because less privacy-oriented individuals in my family can effectively hand out "close enough" data for my genetic information.

Though honestly, DNA is probably the least worst data law enforcement could be using against me. It's reasonably accurate, and it requires they have actual physical evidence in a case they're investigating to compare against. While it's true I could've touched an object shortly before a killer did, it's very far from the dragnet approach of say... querying the identities of all cell phone users in a geographic area at a given time. If my identity pops up in a DNA search, it's reasonable that I will, at the very least, have useful data about the incident at hand.

I would like more protection though for search of these sorts of databases though. Searching a DNA database should require a warrant of some kind. For a judge to look at where the DNA was found and the circumstances of the case, and decide it's reasonable and warranted to conduct the search.

It's the combination that really scares me.

Your cell phone located you in the area at about the right time, and your "close enough" dna was the only dna found on the murder weapon. Slam dunk case.

Don't even really need people to put this evidence together: AI can do it! What can possibly go wrong?

Not sure if anyone's done this but I specifically put a fake name and email when I submitted my DNA test samples to maintain my privacy as much as possible.

I suppose they could track me based on my mailing address, which I didn't spoof, but it's better than nothing.

Who do you think would track you and for what purpose?
The scary thing about DNA and all of its future use is that all of your "privacy" measures are completely irrelevant, as the DNA of whoever is genetically related to you and known will easily lead back to you. And the probability of someone in your extended family being registered in a genetic database will closely approach one within the next decade.
Specifically that's how they found a serial rapist. The rapist hadn't submitted his DNA, but enough of his family had.
I was reading somewhere on a security website that the topsy-turvy thing about that was that the impoverished, who tend to have entire families that are impoverished, are far less likely to ever be in these commercial DNA databases. So the people most likely to commit a crime, are the least likely to have DNA on file. (Of course he went on to point out that the poor are only more likely to commit property offenses, and property offenses statistically require less DNA evidence to solve than personal offenses. So it wouldn't matter much in practice.)

It was an interesting take on the familial DNA issue with commercial DNA databases.

> impoverished, who tend to have entire families that are impoverished, are far less likely to ever be in these commercial DNA databases. So the people most likely to commit a crime

Well if we're going to make overgeneralizations like that, I could guess that the impoverished on average are more honorable than the rich, and less likely to commit crimes. In fact, that honor might even be the cause of their impoverishment. Crime pays and when the richest are so because they're willing to take money from under the table, outright steal it, or "earn" it from selling to addicts, what chance do the people that are trying to earn a living through honest means have? You know there are cemeteries for drug lords where the each tomb is a little mansion?

There's also studies where free money was given away to anyone that wanted to take it. Most looked well off and said they wanted it for purse or whatever taking a ton of bills. Homeless on the other hand would just take a few bucks to buy a hotdog or other necesity, too ashamed to take more.

There's also videos online of people trying to steal packages (the homeowner set up a camera and trap). All looked well off. They didn't look homeless or sketchy in any way. Some were even driving good cars.

Let's not judge people on the amount of money they have.

I don't think they were really making a moral statement saying that crimes are committed by bad people, but that crimes are often committed by people in desperate situations and reduced opportunities, which I think is a reasonable, non-moralistic interpretation of the statistics.
> saying that crimes are committed by bad people

That's different, and probably an objectively true statement unless law is written badly or by a thoroughly corrupt government.

> but that crimes are often committed by people in desperate situations and reduced opportunities

That's what I'm not sure I'd agree with. It's just as logical to think that people (not all) in desperate situations are so because they're restrictive in their actions, taking only opportunities that are in completely in line with their moral ideals. It's also logical to think that people (also not all) that are well off might be so because they took opportunities that might not have been of the highest morality (if not to say that they might have been outright illegal).

> which I think is a reasonable, non-moralistic interpretation of the statistics.

Can you cite those statistics?

What I dislike about that type of thinking is that it leads to prejudice, so I'm not sure it's "non-moralistic" in any case.

Anyway even if there is a location where most poor are bad and most rich are good, this seems to be very location and culture dependent. I don't think it's a claim that can be shown to be universally true.

I'm slightly confused by this comment. If it was important for you to not have traceable DNA, why wouldn't you just not submit it?
Presumably OP was interested in the details of the DNA assay without being integrated accurately into a larger corpus of genetic data.
I did this for 23andMe. If I were truly privacy conscious I wouldn't have submitted at all, but this at least provides a small layer of defense in some minor threat models. What was funny, though, was when I did the search for relatives feature after, it listed a bunch of people with my fairly unique real last name, so oh well. Neat bit of evidence that the sequencing worked anyway.
This is like saying, "I looked up information on how to make bombs, but I put my browser in incognito mode. Yeah, the FBI can still get my browsing history from my ISP, but it's better than nothing."

In both cases, it's not better than nothing. It's exactly nothing.

People get so worked up over this stuff, for no good reason. DNA evidence has solved so many crimes and done so much good. I think every single US citizen should have their DNA on file. Period.
I think the ability to solve a criminal act via the technique described is perfectly valid. It's quite ingenious in fact.

What GREATLY concerns me is the thought of what happens when one of these companies gets down on their luck and sells all this data to an insurance company. Can you imagine what would happen to your rates and coverage if it comes back that your family has, say, a predisposition towards diabetes or cancer? There needs to be legislation enacted to strictly control how this information gets used. Furthermore an individual should be allowed to control how that information is shared and be able to opt out of any scenarios they are not comfortable with. Law enforcement scenarios could be exempt but tightly regulated to avoid abuse.

[edit] - its been pointed out that insurance co's can't do what I am describing (which is awesome btw). I think my point stands about the implications of this data being monetized to our detriment.

Current US law is that the insurance company can consider age and tobacco use.

That's it.

Good chance it won't last though.

I did not know that. :-)

You get what I mean though. This is valuable data that can be exploited.

> Can you imagine what would happen to your rates and coverage if it comes back that your family has, say, a predisposition towards diabetes or cancer? There needs to be legislation enacted to strictly control how this information gets used. __Furthermore an individual should be allowed to control how that information is shared and be able to opt out of any scenarios they are not comfortable with.__

Emphasis mine... because I wonder, even if individuals opt into various things, what about their descendants? It's not too hard to imagine a family tree with people opted-in but a descendant who doesn't want their information to be available -- but the general risks/predispositions would be. I haven't seen anyone talking about the issue of privacy with respect to inheritance yet.

Kind of like how you can “choose” (in theory) not to give any information to Facebook et al, but if your friends and family share their contact lists and email boxes, they get plenty of info on you anyway.
Universal health care
Doesn't solve the problem. Banks will care about this, employers and even landlords would care about this stuff.
Of course they will care, the point it to find solution to make it doesn't matter however they use the information. In case of universal health care, it doesn't matter if insurance company exclude you for pre existing condition or whatever, universal health care will cover you.
>Can you imagine what would happen to your rates and coverage if it comes back that your family has, say, a predisposition towards diabetes or cancer?...

Too late.

They already know this. The DNA would only verify the findings they've teased out of the mountains of data they have on us presently.

What you describe is the Genetic Information Nondiscrimination Act (GINA)

http://www.ginahelp.org/GINAhelp.pdf

Just a note that GINA doesn't apply to life insurance and mainly covers health insurance and discrimination from employers.

> 
They
 do
 not 
apply 
to 
life,
 long‐term 
care, 
or 
disability
insurance.

(comment deleted)
What happens when your sample is contaminated with a criminal's (or future criminal)? What if a clean sample generates a false positive match to something undesirable? What do you do when your erroneous genetic records close off doors or sends you through the criminal justice system? You will be powerless to fix it.

You are at the complete mercy of lab techs not screwing anything up. It's safer not to put personally identifying information into a database.

Remember Annie Dookhan!
I had to look her up: https://en.wikipedia.org/wiki/Annie_Dookhan

She was a chemist at a crime lab who falsified large amounts of evidence.

This is pretty shocking.

> On December 17, Dookhan was formally arraigned on 27 charges—17 counts of obstruction of justice, eight counts of tampering with evidence and one count each of perjury and falsification of records.

> [The judge presiding over Dookhan's case] wrote, "Innocent persons were incarcerated, guilty persons have been released to further endanger the public, millions and millions of public dollars are being expended to deal with the chaos Ms. Dookhan created, and the integrity of the criminal justice system has been shaken to the core."

All of that resulted in a sentence of only three years (Nov 2013 to April 2016).

Wow, I did learn something today.

"In May 2015, the Massachusetts Supreme Judicial Court ruled that defendants whose convictions on drug charges were based on evidence potentially tainted by Dookhan can pursue retrials without having to face more charges or tougher sentences."

Wouldn't it be more appropriate to mass release everybody, then have the prosecution retry who they think they can convict?

>>what happens when one of these companies gets down on their luck and sells all this data to an insurance company

If it's not illegal already it can be made. No serious company will even wonder about the what-ifs then.

>I think the ability to solve a criminal act via the technique described is perfectly valid.

This would be true if laws were divine instruments of justice rather than poorly crafted, poorly-thought out instruments of social control crafted by the wealthy and powerful to control the poor and powerless. For example, we as a society are far, far better off not having "perfect" law enforcement in regards to drug laws. Solving every drug crime and prosecuting every "transgressor" would result in tens or hundreds of millions more destroyed lives and families, with absolutely no positive benefits to society. There are very good reasons to ensure that the power wielded by agents of the state is finite and limited, even (or especially) if it means that many "crimes" are not solved.

We are absolutely on the same page about the draconian nature of US drug laws.

I think there can be a middle ground. The GTK being caught and brought to justice for instance. That is a concrete example of the good that can come out of this. A monster like that is an affront to society and he needed to be caught if only to show that there can be some justice in this world.

Someone posted a comment above mentioning Annie Dookhan which is an example of how wrong this can go. I think any rational person would be terrified of something like that happening to them. I certainly am. I don't think a blanket ban is the answer however. That would result in many egregious injustices to go unpunished as well. Any application needs to have a robust set of checks and balances in place. That really goes for any law enforcement investigative action though. No matter what we do there will be issues. I don't there is anything to gain on either end by trying to make this a zero sum game.

> What GREATLY concerns me is the thought of what happens when one of these companies gets down on their luck and sells all this data to an insurance company.

What reasons are there that a company has to go down on their luck to do this?

This article is so FUD that it hurts.

Honestly, yes, I want law enforcement to use DNA among other evidence they collect. I do not live in fear that law enforcement is secretly hunting me. If you do, speak to someone about your paranoia. My name and address have been on my public resume and my website WHOIS for over 5 years, and I haven't experienced any negative repercussions whatsoever.

I must not fear.

Fear is the mind-killer.

Fear is the little-death that brings total obliteration.

I will face my fear.

I will permit it to pass over me and through me.

And when it has gone past I will turn the inner eye to see its path.

Where the fear has gone there will be nothing. Only I will remain.

DNA is vastly different from your name and address, which many would consider public information. DNA is much more identifying and private.
Yes, and? Do you expect police to come to your house because you're a carrier for Cystic Fibrosis?

Please state your argument in full. Statements like this are just further FUD.

How would you feel about having a taser in a collar at the base of yours (and everyone's) neck? It only gets turned on by the authorities if they see a crime in progress. It's perfectly harmless otherwise, right? If you're not doing anything wrong, what's to fear?

In the case of markers like this, it's pretty easy to correlate by race or other heritable factors criminal behavior. If there's no harm in it, when a crime goes down in an area it seems reasonable to go and just round up all the people with rough genetic markers matching, right?

It might even make sense to keep them in groups organized by genetic markers and race, since that should speed up policework dramatically.

And later, if you want to procreate with somebody, maybe you should make sure their genotype won't mark your kids for poor treatment by the authorities--after all, accidents _do_ happen, and the labwork is never 100%, so better safe than sorry.

Can you see why this makes people antsy?

YMMV, but I consider my residence address (or, generalizing - my live geolocation) way more sensitive than my photo, fingerprints or DNA.

Personally, I have to trust someone much less to send them a photo, than an address of residence.

I'm not exactly sure what can be more undesirable - being recognized or being physically interacted with. However, my first impression when I started to reply to your comment was that it's the latter.

Police states exist. It's a genuine possibility that your "nothing to fear" argument ignores.
I took a course in college where we studied the basic process of DNA amplification. We were just learning at a theoretical level, but from what I understand almost anyone with access to a university biology lab and the associated training could easily create copies of a DNA sample. Doesn't that mean that it would be relatively easy for something like an intelligence organization to frame someone for a crime?
I guess they could point the suspicion to someone innocent, but I would be surprised if law enforcement didn't validate the DNA before sentencing a crime.
The cops have no incentive to run confirmations, for leadership to be re-elected they must appear "tough on crime" or something. There are just no consequences for false positives.

In Orlando the police arrested a fellow after a field test indicated that a box of donuts was coated in meth: https://www.orlandosentinel.com/news/breaking-news/os-krispy...

In North Carolina the police arrested several men after a field test that identified sugar as fentanyl: https://wtkr.com/2018/08/29/2m-worth-of-fentanyl-seized-in-b...

It's well known that that sort of test isn't specific for anything and merely gives law enforcement a pretext to arrest a nogoodnik they deem guilty. Why would anyone trust them to act different with DNA? The police's business is arrests, matches are good for arrests, confirmatory tests are bad for arrests, hence no checks.

It's not impossible that it really was fentanyl and some dirty cops stole it.
Your DNA and fingerprints can also be extracted from your trash.
I'm more concerned about the impact on forensics, how these large banks of information will inevitably leak and compound upon developments in the creation of artificial DNA.

DNA evidence is considered infallible within the justice system, should it be? Here's a paper from 2009. [1] Around the same time the paper was published, the author was quoted in a NYT article [2] (again, nearly a decade ago) saying, "“You can just engineer a crime scene, any biology undergraduate could perform this.”" Think about how much this field developed within the past decade. Here [3] is another paper from the end of last year about looking at methylation signatures on DNA to profile the person it belongs to. Imagine the possibility of finding badly degraded DNA at a crime scene and still being able to say, "whoever did this, they have this skin color, they're within this age range, they're a smoker, and they have frequent exposure to this particular industrial chemical." Minority report wasn't imaginative enough.

I'm concerned about this being used politically, I think it's possible with enough money and motivation. I think our courts are too addicted to open-and-shut convictions to reevaluate the validity of DNA evidence going forward.

[1]: https://doi.org/10.1016/j.fsigen.2009.06.009

[2]: https://www.nytimes.com/2009/08/18/science/18dna.html

[3]: https://doi.org/10.1186/s13059-017-1373-1

(comment deleted)
> our courts are too addicted to open-and-shut convictions to reevaluate the validity of DNA evidence going forward

Have you ever served on a jury? These questions are deeply contemplated in selection, presentation and deliberation. If the evidence could have been falsified, a jury neutral to that argument will hear the defense’s case.

Having had a jury, there are way too many things wrong with that thought to know where to begin.

Let's start with jury selection. If a venireman (that's somebody in the jury pool) has anything objectionable known about themselves to either side, one or both sides will strike them. This leaves juries with muppets (in the sense of people who will never themselves make the history books) and the occasional stealth.

Juries aren't neutral to anything. The expert the state hires goes up and says, yup, DNA matches. Defendant probably doesn't get an expert even if he's certain he needs one. And that's only if the judge allows it, which is often not the case.

Half of the jurors are already pissed off at the defendant because that's how people are- he decided to protest his innocence and now I have to miss a week of work for less than minimum wage.

DNA unfortunately is all too often junk science, because in the instances where the forensic lab technician isn't adulterating the tests intentionally, they can be doing so unintentionally. The DNA you're not looking for can often get in the way. This is without Chimerism, and other recently discovered genetic oddities muddying things up further.

One need only look at Milgram's experiment to know how most jury trials go.

What are you joking?

They strike anyone with a brain or domain-level knowledge (until they run out of strikes).

You never see the totality of the evidence; just what the judge "allows in" (that isn't prejudicial to the prosecutors case).

You don't know what the punishments are; half the time you aren't instructed on lesser offences; and most of the time you never even get to see the actual laws, or more importantly, case law.

if you think sitting on a jury (as it stands today) qualifies you to render any sort of judgement on the "evidence", you are woefully mistaken.

You are a puppet. A fop. Anything but "the last bulwark against state tyranny".

The appearance of justice, is the only thing Courts give a shit about. You are part of that appearance. A show, for the uneducated to think they still have a functioning legal system.

No. DNA has already been shown to be bogus. Just look up any of the stories of "transfer".

And when you're thinking, well what about rape? Consider that there are documented cases of women taking used condoms and shoving semen up their cunts (case was not a rape case; but fertility rights).

DNA, doesn't prove shit. You should stop, treating it, like it does. Particularly when countless "law enforcement" labs have been proven to be falsifying evidence.