Across the United States, and increasingly internationally, detectives are closing cases that have long since grown cold using investigative genealogy. Killers and other predators who have managed to escape justice due to a lack of leads are now finding themselves behind bars.
Investigative genealogy relies heavily on open-source DNA databases, such as GEDMatch, where users upload their DNA raw data files. Originally intended for use by genealogy hobbyists, the site has been a gold mine for investigators. Once they upload DNA profiles of suspects taken from crime scenes, they are then able to search for close familial matches and try to build the suspect’s family tree. Once the tree has been built, investigators are able to uncover the identity of the suspect, giving them a lead in the case.
Since 2018, over 55 cases have been closed, and several John and Jane Does have been identified using investigative genealogy techniques, but the future of the technique is unclear.
On November 17, 2018, a while she was practicing the organ in a church meetinghouse. CeCe Moore, an investigative genealogist, was asked to assist with the case. Knowing that using the GEDMatch database to solve an assault case would violate their terms of service, she initially declined. With express permission from Curtis Rogers, founder of GEDMatch, investigators were allowed to use the database to identify the attacker.
Facing public backlash for violating his website’s terms of service, Rogers quickly updated GEDMatch policy to require users to opt-in to allowing their profiles to be used for investigative purposes. Where there were once almost 2 million searchable profiles, 20,000 remained as of June 9, 2019.
Though users continue to opt-in, where does this leave the future of investigative genealogy? CeCe Moore told CNN, “We can be sure there are hundreds of cases that would have been solved in the coming months or year that very well won’t be now. People will die. It sounds dramatic, but it’s actually true.”
What does the future hold for forensic labs interested in investigative genealogy? No one knows for sure, but we interviewed a panel of experts to get their opinion.
GEDmatch has recently changed all uploaded profiles in their database to a default opt out for searches by law enforcement. How do you see this impacting the number of cases solved using genetic genealogy in the future?
Colleen Fitzpatrick and Margaret Press (DNA Doe Project):
I think the number of cases solved will slow down in the near future, but this will only be temporary. As GEDmatch recovers and as uploads to FTDNA increase, the number of solves will pick up again. However I think that the rate of solves was going to drop anyway as the easy cases were exhausted.
The situation before the change in GEDmatch was much the same as when the State of California first automated fingerprint identification. At first, there were a large number of fingerprint identifications made that were “waiting to happen”. I refer to these as Oh My God (OMG) cases. But once the easy cases were exhausted, the number of IDs leveled at some steady rate, that is probably gradually increasing due to the submission of new cases that is causing the database to gradually increase in size.
The change in GEDmatch has disrupted this process, but as GEDmatch’s size gradually increases, the number of solves will eventually reach the steady state it would have had had there been no disruption. But the OMG era is probably over.
Diahan Southard (Your DNA Guide):
I think our current situation is much like how life was when the first digital music devices became available. If you had one, it was difficult to get music onto it, it just wasn’t readily available. That is how are databases are now. Small, clunky, hard to access and use. But just like digital music has completely become the norm for nearly everyone, this tool of genetic genealogy will grow and become easier to access and use for law enforcement.
Is there an upside to the GEDmatch change? For example, does this alleviate privacy concerns?
The upside of the GEDmatch auto-opt-out is that it made the international news. Although an unknown number of individuals took down their kits, the opt-in recovery has had informative patterns. A startling number opting in, in excess of 25%, are new transfers to GEDmatch from direct-to-consumer test takers who had not sent their kits previously. The additional media coverage had a positive effect on the two GEDmatch owners, who had not created their site with law enforcement uses in mind.
The change did nothing to alleviate privacy concerns, it heightened them, particularly in communities like followers of a few genealogical bloggers. The tiny numbers of African Americans and EU residents who have opted in to law enforcement permissions show that the discussion surrounding the change alarmed people who have good reason to fear government misbehavior. In real life, if you are risk averse to real dangers you should not test, let alone post. Since its inception, GEDmatch said this very clearly. Changes to that concise statement have done them considerable harm. Law enforcement searches, usually done by civilian genealogists, have been non-intrusive, exonerated countless people, and anecdotally deterred criminals who know they will be caught.
Allison Nunes (DNA Labs International):
I do think there is an upside to ethical and privacy concerns, and this policy provides more transparency for the general public. Logistically there is a huge benefit to GEDmatch because it also eliminates the potential for individuals filing to have their profiles removed on a one-by-one basis. Enacting this policy change now, gets rid of a potential issue early on in the process which should make for a smoother future for forensic genealogy.
Should a reduction in the number of profiles available for searches discourage labs that might be considering implementation of investigative genealogy techniques?
Implementing investigative genealogy on a lab by lab basis is not as simple as it sounds. Laboratories that are considering investigative genealogy should first assess how many cases it could potentially benefit, and from there determine the best way to meet those needs, whether it be complete outsourcing or outsourcing of just the single nucleotide polymorphism (SNP), whole genome sequencing (WGS) or Y-STR testing.
Absolutely not. After all, you don’t need a million samples to solve your case. You only need a few. And perhaps the few you need are already there in the database.