First, let’s establish that I am not an attorney. I talk to lots of attorneys and even attend law seminars from time-to-time, but I have never studied law in any meaningful way. My post on this matter is because I have a gut feeling the use of DNA searching in the case of the Golden State Killer is probably new territory.
Second, let’s establish that though I’m not a geneticist, I am experienced enough in genetic genealogy that I can tell you mainstream media reporting on DNA testing is as terrible as it is on guns. Most reporters know absolutely nothing about the types of DNA that can be tested in the consumer market, how matching happens, what the policies on use are with those sites, and the limitations of DNA. DNA doesn’t lie, but it’s sure as hell misinterpreted – often. Add to the problem of reporters not knowing anything about it the fact that most of the people in front of the cameras from the police department and the district attorney also know nothing about it, and you’ve got a recipe for terrible reporting.
Based on some of the most detailed reporting that didn’t come out until late last night and this morning, the investigators did not use a commercial DNA testing company you commonly see advertised. In the US, the big ones are Ancestry.com, 23andMe, FamilyTreeDNA, and MyHeritage. Three out of the four confirmed in statements when the media started getting things wrong that they in no way cooperated with any criminal investigation with their DNA databases. While the fourth one didn’t issue a statement, the police didn’t use their private databases, either. All of them have pretty solid policies not cooperating without a warrant.
However, because these companies all offer different tools, many of which are useful in their own ways to us genealogists, but you can’t compare to someone who tested at another site, some folks built a website called GedMatch that’s sort of like a public open source comparison tool. It’s database is made up of people who have taken their raw DNA data from other companies that run the tests and upload them to be compared to each other. Why? Well, in cases like my grandfather’s, he only tested at Family Tree DNA before he died. If a great match tests at Ancestry, we can’t compare the two kits except at Gedmatch. (Technically there are a few alternatives now, but there weren’t for a long time. Gedmatch became the place to do it.)
The police created a raw data format that looked like one of the major testing company’s raw data forms based on the DNA sample recovered from a crime scene. Using that form, they uploaded it to Gedmatch as though it was a sample from one of the major testing companies. There’s nothing in Gedmatch’s terms that says you can’t “spoof” a major company format with data obtained elsewhere. The entire purpose of the site is to compare across testing platforms.
BUT, where things get interesting is that when you upload a kit at Gedmatch, you must answer a question:
Please acknowledge that any sample you submit is either your DNA or the DNA of a person for whom you are a legal guardian or have obtained authorization to upload their DNA to GEDmatch: Yes or No
Now, the officer who uploaded it 1) is not the source of the DNA, 2) is not the legal guardian of the suspect, and 3) does not have the authorization of the DNA sample provider to upload. Yet, s/he uploaded it anyway by directly misrepresenting his/her status as the legal representative with permission of the source of the sample.
Only with that misrepresentation did they find enough information to focus in on one subject and follow him around to find a source of publicly disposed DNA to run a direct test against the original sample using the traditional law enforcement test (CODIS).
Is the misrepresentation of who has permission to use a website under specific terms a violation on the part of police? I think it’s a reasonable argument to make, though I don’t know how much precedent we have in this area of law because we’re talking technology that most previous cases don’t really cover.
Now, GedMatch was also recently used to solve a Jane Doe victim status. I tend to believe that there’s a more solid case that police are legal representatives of an unidentified deceased victim in their “custody” than they are over a possibly still living suspect’s DNA sample. That said, any law enforcement use of the database makes people, understandably, uneasy. There have been calls for years for law enforcement to consider an opt in GedMatch-like database in order to solve John/Jane Doe cases with clear terms and conditions for those who opt in. Plenty of people are willing to help in those cases, and even in searches for criminals. But the authorities didn’t give people that option, they held themselves out as representatives of criminal suspects to gain access to a database.
In the context of guns, Sebastian & I figured it is most closely represented with this type of analogy:
A plainclothes cop shows up to your gated, private gun club and tells a board member that he has a document attesting that he’s a representative & there with permission of your club’s Membership Chairman and is just looking some things up on his behalf without ever disclosing that he’s law enforcement. He is allowed into the gated club only based on this form that says he really is representing your Membership Chairman, even though the officer really plans to investigate your Membership Chairman on a case not related to your club or its business. When he comes in, the officer looks through the guest log and sees a name he recognizes as similar to a felon known to him. He uses the information on the guest log to start following around your Membership Chairman to see if he’s allowing a felon to use his firearms. When it turns out to be the case, the cop sweeps in to make an arrest. Now technically the arrest is based on what he witnessed. But he never would have been following him if not for what he saw when he entered the gated property by falsifying documents that claim he was the Membership Chairman’s representative and was there with his permission. Is this a fruit of the poison tree problem?
I don’t know exactly how this will all play out, and I don’t know enough law to know if this is absolutely a violation. What I do know is that it violates ethics of the genetic genealogy community. While I understand that GedMatch doesn’t really have the same “gates” restricting access to its database as the commercial companies, the fact that they aren’t addressing the very real issue of how the cops represented themselves in order to be matched against others bothers me. That question in their upload process either means something or it doesn’t, and GedMatch isn’t taking it seriously. I made our kits private in the meantime.
As a final thought on DNA & crimes. I think it’s important to remember that as much as guns are misrepresented and therefore misunderstood, so is DNA.
DNA alone can tell you remarkably little without extra context and information. It doesn’t lie, but it is sure as hell misunderstood and treated as inappropriately damning. Consider that California was going to put a man away for life for a murder he absolutely did not commit based on the officer’s lack of detailed analysis in considering how DNA might have appeared on a victim. I mean we all “know” that if you find someone else’s DNA under a victim’s nails and they struggled, then that must be your assailant. It’s what tv tells us! Except what if the EMT called to the scene reuses a pulse oximeter on the victim that was used hours before on a man he took to the hospital? That was an important detail that nearly cost a man his life. Even in genealogy, something as obvious as a parent/child match? If you don’t know whether that parent has an identical twin, then you don’t have all of the information you need to make a sound conclusion.
It’s complex. And that’s why the media will f*ck it up almost every time they report on it – just like guns and gun owners.
UPDATE: The NYT has a decent article on the ethics & possible legality issue. A law professor who professes to know about DNA searches says there may be questions about the legality of the evidence. I’m going to assume that’s why they went after the other samples so that any warrants were technically issued on those grounds rather than the legally dubious GedMatch grounds.
Courts have allowed in other cases that the police are sometimes allowed to lie (and even commit other crimes) in pursuit of criminal activity. I wonder if this will fall in that area.
Sebastian found a case where a police officer told someone he had a warrant – he didn’t – in order to obtain permission to search. The courts ruled that such a blatant misrepresentation in order to convince people to go along with the investigation is a violation.
Thank you.
I saw your original comments at Miguel’s site, and I am glad you took the time to expand and explain this further.
Good post. Just for the record of DNA forensic f’ups, here’s another one. Article is from 2009, but it details how police were chasing a mystery female killer for 15 years. She was killing people across Europe, relentless. Turns out the swabs were contaminated with the DNA of the lab tech. So, yeah.
https://www.independent.co.uk/news/world/europe/dna-blunder-creates-phantom-serial-killer-1655375.html
It never should have taken that long to re-examine the chain of custody of the kits once they saw no consistency in the locations, crimes, etc.
Was it the lab tech?
Or was it someone involved in manufacturing and packaging of the “sterile and forensic quality” swabs?
Don’t they have DNA samples from their own techs on file, just to quickly identify and eliminate such errors? Just like having the finger print results of people with normal approved access to the scene, so they can be eliminated from the search for outsiders? Of course, admitting that the sample was contaminated in the lab, by the lab techs themselves, is a BIG RED FLAG.
The more I read about the science of forensics, the more I think much of it is pseudoscience. “Bite Mark Science.” “Lead Alloy Composition.”
How many labs have had friendly and helpful techs overstate the quantity or purity of the drugs just to help out the police and prosecutors? I can recall cases in Massachusetts and Minnesota off the top of my head. There was also a big scandal at the FBI Labs themselves at the turn of the millennium.
I don’t hate the police, labs, and prosecutors, it is just that their own incompetence, corrupt, and worse, ass covering behaviors had made me damned wary of them.
In the linked case, it wasn’t even contaminated when it got to the lab, it was contaminated before it ever arrived to the police to use. It was in production, so they never thought to run a check against the people producing the kits that were purchased. I can see the first few cases not setting off alarm bells, but once they had a wider geographic area and no other connected trends, they should have realized along the line that all of the kits came from the same source and had the tests run.
There’s another case, and it’s more similar to this one–AND, it’s the reason that other genealogy sites now require warrants like they should have in the first place–http://www.theadvocate.com/new_orleans/news/article_1b3a3f96-d574-59e0-9c6a-c3c7c0d2f166.html
Even without looking at your link, I know that case. That case was a different type of DNA, and I will admit that I have no idea what the policies were on the old public Sorenson site to be run against their matching database. Because I started this journey out in autosomal DNA in 2013, I didn’t really know about most of the older Y-DNA sites and I don’t know how much the police had to do to represent themselves as a potential representative of a suspect or if they could just type the markers in without any claim of being connected to the sample “provider.”
That said, one of the other issues I have is that there’s no requirement that they consult with qualified people. (To be fair, there’s not much in the way of qualified people, but the person they consulted on this case didn’t know jack about DNA, IMO.) I’m most knowledgable about autosomal DNA, but I dabble in Y-DNA. (I only manage/access about 5 Y-DNA kits compared to double digits of autosomal.) But I’ve talked to enough Y-DNA research project administrators and have enough experience in those handful of kits to know that the number of markers they used isn’t remotely acceptable for the kinds of accusations they were leveling. Yet t
Add in that the rest of their case for getting approval from a judge was misrepresenting what the DNA said and the fact that the man they wanted to target once drove through the state. Any officer, DA, or judge who makes a case for a suspect based on the evidence that they once drove through a state needs to be out of a job yesterday. Even when paired with completely misrepresented DNA “evidence,” this type of DNA can have a perfect match from hundreds to a potentially thousands of years ago.
Thanks Bitter, for your in-depth reporting and analysis. It just figures the Press would screw up reporting on something important that deals with technology. Press incompetence is going to have worse and worse consequences for our nation as technological public policy issues become ever more important.
We really do live in an idiocracy, don’t we? Only it isn’t as the Press and our Power-Elites assume, an idiocracy of the masses, in reality they are the idiots!
It would be useful if a learned member of the legal commnity – and by that I mean “extremely highly skilled in understanding and interpreting DNA results” – could chime in.
I suspect the legal community will have to engage support personnel who have complete understanding of the complexities of DNA, and that can, and will, easily lead to dueling experts. For which there is no substitute when confronted by “confirming DNA “evidence” from the authorities. So far, DNA evidence is accepted as gospel, and often that’s not the case.
The bigger challenges, probably, will be adequate certification for experts and the ability of those experts to clearly explain the complexities of, and differences in, DNA results that juries can comprehend.
Dueling experts in both law and science usually result in better outcomes and understanding, so this is something I endorse! Unfortunately, the typical person in a jury pool won’t and most will probably decide based on which expert they liked more rather than the substance of what they say.
There is an attorney in the genetic genealogy blogosphere, and she usually posts about DNA on Sundays. I’m sure her site will be inundated if it’s not already. But I will say that I don’t know how specialized she is in criminal law, especially current precedent on 4th amendment issues. There’s another attorney in this area, and while I haven’t had a chance to find exactly what he has said about it, I understand he may have misgivings about this as well. Granted, since both were involved in drafting the generally accepted community standards of ethics in genetic genealogy, I’m not surprised. Legal or not, this was highly unethical for them to hold themselves out as uploading to GedMatch with the suspect’s permission.
MSM reporting sucks on pretty much everything, not just guns and DNA. I have been present at hundreds of events covered by the media and not one single time did they get it completely correct. Some of the errors were minor or technical but large numbers of the reports left me wondering whether they were at the same place I was.
No doubt. I was a fireman for 30 years and was frequently interviewed. Every single time I was either misquoted or some fact was misstated. Sometimes only slightly, but not once did the reporter ever nail it.
Michael Crichton observed this, and then pointed out that after seeing how the media botches up something we know a lot about, we go on to accept, with a high level of credibility, what they have to say on things we don’t personally know about. He called this the “Murray Gell-Mann Amnesia”.
Robert Heinlein made (an extremely tangential) observation on the same subject in an essay collected in Expanded Universe. Written sometime in the early Sixties, IIRC.
Long time reader, first time commenter. I am a criminal defense attorney, so I may actually be able to ad something useful to this one.
The issue about whether this is permissible is going to come down to a question of standing, i.e., whose rights were violated? For instance, if the police, through force or fraud, come into a person’s house and seize evidence when not otherwise legally able to do so, that person could challenge that search in court, potentially getting an order from a judge excluding that evidence, as well as any other evidence derived from it, from the proceeding. This is called the exclusionary rule. The idea is that the mere threat of a trespass lawsuit from a convicted criminal is not really a deterrent to illegal behavior by law enforcement, but excluding evidence is. This is an extreme remedy, however, as it often compels courts to let someone known to be guilty go free. So, the courts have imposed limits on its application. One of those limits is that, generally speaking, only the person whose rights have actually been violated can successfully exclude the evidence.
For example, let’s say that a friend and I go commit a burglary, and we videotape the whole thing. Afterward, we go our separate ways, and he takes the videotape home with him. Sometime later, the police barge into his house without a warrant, illegally search it, and find the tape. Although it affects me, he is the one with a right to privacy in his house, not me. He may well be successful in challenging the search, but the government will still be able to use that tape against me.
The same theory applies to many aspects of modern life. Use Gmail? Your messages are in Google’s possession, not yours. You have no enforceable Constitutional right to privacy in them; though, thankfully, Congress has passed laws restricting what police can do in this arena. To my knowledge, though, no such legislation is going to protect you when you have willfully disclosed your genome to a third party, particularly when you know it is going to be made available to others. It’s definitely not going to protect you when they only get information about your relatives, rather than you.
Take care when turning over your property to a third party. In the Age of Information, this has become more important than ever.
That’s kind of what Sebastian ultimately thought the issue would come down to. Technically, it’s GedMatch and perhaps the DNA matches to the sample who were “harmed” in the violation of the site’s terms of upload. It’s doubtful any of them will object.
In fact, it’s GedMatch’s refusal to acknowledge the violation of the upload agreement at all that has me pissed off at them. They don’t have to go fighting anything in court, but it’s reasonable to acknowledge that a) there is an agreement that everyone must “sign,” and b) there’s a reasonable legal debate over whether or not the police lied when they uploaded the kit.
I can’t help but wonder, though, what GedMatch’s recourse is. They can delete the DNA info that the police entered into the system, I suppose — but that’s basically closing the barn door after the horses have left. Similarly, I suppose they could sue the police on behalf of the person whose DNA they submitted — but then, how was GedMatch harmed by adding this to the database?
I also suppose that the person whose DNA was submitted could also sue…but strictly speaking, while the DNA sample was taken without his permission, it was taken well within the parameters of a police investigation — so I fail to see how the person has any right to complain about how the police decide to use it.
The more I think about it, though, I can’t help but wonder: is it really too much of a burden to ask the police to get a warrant for searching a public database, so that even if they answer Yes to “Please acknowledge that any sample you submit is either your DNA or the DNA of a person for whom you are a legal guardian or have obtained authorization to upload their DNA to GEDmatch: Yes or No”, they do so with legal authority?
Cops won’t get warrants unless they absolutely have to; as a couple of lawyer acquaintances made sure to tell me in discussions on a different topic.
Keeping control of your own DNA and genetic information to me is critical. My brother had a genetic flaw that affected his heart and hand . It was in vitro mutation. They(he and his wife) wanted genetic testing to confirm, but thought it would be released to insurance companies as a reason to deny coverage. So they did not.
This case immediately raised concerns to me of the confidentiality of the DNA database by commercial vendors Made me even more disinclined to have this information in another’s control.
They didn’t use any of the commercial site databases. They used what’s really an open source-style site that people upload into.
That said, one important protection you have in the DTC field, there’s no chain of custody to prove you did or did not take the test. If you have matching turned on, then depending on your age and how much may have been published in obituaries and other family history sources (like Facebook, actually), your genetic matches may be able to figure out who you are. But there are plenty of things that can trip you up, too. I’ve had extremely close matches I can’t place due to a number of factors and the best I can do is say, “well, I know they are a grandchild of these two people.” But because paperwork doesn’t always match DNA, that doesn’t provide much to go on in some cases.
Did you see this story? The search (which used only a part of the killer’s genetic profile) originally identified the wrong person?
https://pjmedia.com/instapundit/295386/
I wonder how many other people and families were investigated and eliminated as part of the manhunt?
Yes, I did hear about that portion of the search. From what I’ve read about that, they tried the same method as they did in the case that WallPhone linked above. They followed clues from a public Y-chromosome database rather than an autosomal database, which is what Gedmatch is. (Two totally different types of tests, types of DNA, and portions of your family tree.)
Y-DNA is going to burn them every time they to use it when they are searching blindly. Of the handful of kits I manage, I have one where paternal 1st cousins have a genetic distance of 3, which is something like a less than .5% chance. (I’d have to log in to see the stats exactly, but it’s ridiculously rare. Due to mutations up the tree to a 4th cousin & autosomal matches with others closely related to the 1st cousin, we know it’s accurate.) I’ve also seen perfect matches with people who probably don’t have a common ancestor in the last 400 years.
I don’t know how to make them learn. But it definitely makes me far less likely to work with others in genetic genealogy because I know the cops are getting it wrong repeatedly, which is quite sad because that’s how we make breakthroughs. Using DNA, I’ve narrowed down a search that has baffled my family members for a good 70 years to two possible sets of parents. I’m not quite sure DNA can help us figure out which of the two guys are fathers since they were brothers, and I’m pretty sure their wives were related given the level of endogamy in the family. Plus, the courthouse burned. Twice. Still, even with burned courthouses, common and practically unsearchable names, DNA has still led me to more family.
Teh update doesn’t make me feel a lot better, because parallel construction is an even worse issue.
“We know what we did was wrong, so we’re going to pretend we did it the right way in the first place.”