The rising and controversial trend of catching killers using genetic genealogy

Visit any police website across the UK and you will be met with rows of faces – each one belonging to the victim of a crime that has yet to be solved. As of March 2022, the last full year for which we have statistics, only 5.6 per cent of crimes in the UK secured a conviction. At the same time, Home Office figures show sex offences hit a record high, homicides were up 25 per cent, and 2.4 million cases were closed due to “evidential difficulties”, without police ever identifying a suspect.

There is a relatively nascent technology that has potential to reverse such trends, called forensic or investigative genetic genealogy (IGG). It’s been both lauded a success and had its legality called into question, and as officials decide its fate, we look at where the battle lines have been drawn.

Since the first at-home DNA kits hit the market in 2007, more than 26 million people have uploaded their DNA online. The early tests revealed interesting but not particularly useful information about a person’s geographical heritage. As they have evolved, the tests have unlocked more actionable insights, to the point where today they reveal which genetic markers are responsible for your eye and hair colour, or your predisposition to health conditions –everything from the genetic likelihood of you having excess ear wax to your cancer risk. 

As the number of people taking these tests has risen, it’s now possible to compare your results to millions of others to find relatives, plot shared ancestors, and fill in gaps on family trees. Outside of the consumer use cases, increased understanding and access to DNA has allowed researchers to trace ancient lineages and identify remains. When a body, believed to be King Richard III, was found under a car park in Leicester, researchers compared DNA from the remains to someone believed to be Richard’s distant nephew to confirm his identity.

It was only a matter of time, then, until the idea of identifying modern remains, and catching criminals using genetic genealogical techniques, would pique the interest of law-enforcement bodies.

In 2018, one of the most notorious serial killers in recent history was caught using genetic genealogy. The so-called Golden State Killer was responsible for at least 13 murders, 50 rapes and more than 100 burglaries in California during the 1970s and ’80s. Law enforcement had his DNA, but without a match on their internal databases, the killer’s identity remained a mystery for decades.
That was until investigators ran the DNA through a public database called GEDmatch. This ultimately led them to Joseph DeAngelo (pictured), a former police officer, whose DNA was matched to the samples recovered from the crime scenes. He was arrested and charged with multiple counts of murder. Overnight, this single case thrust genetic genealogy and GEDmatch into the global spotlight.

Founded in 2010 by genealogy hobbyists Curtis Rogers and John Olson, GEDmatch allows users of at-home DNA test kits to upload their results and find matches. The database goes beyond the services offered by genealogy sites because it’s public, and it allows people to compare DNA from multiple databases, not just the service they’re subscribed to. Stats suggest the number of crimes solved worldwide using genetic genealogy exceeds 400 (almost exclusively in the United States), and 120 of these have been attributed to GEDmatch.

In the past year alone, the 52-year-old unsolved murder of a Vermont woman, Rita Curran, was solved using genetic genealogy. The parents of a baby dumped in a bin in Mississippi in 1992 were genetically identified and charged. The remains of a woman found in Oregon in 2019 – dubbed the Sweet Home Skeleton – were identified, while the murders of Nancy Bennallack (stabbed in 1970), Anna Kane, and Stacey Lynn (both strangled in 1988), were all solved using DNA.

Yet it’s not just investigators in old and cold cases that are leaning on IGG. Bryan Christopher Kohberger was arrested in January, charged with murdering students at the University of Idaho in November 2022. Reports claim he, too, was identified using genetic genealogy, although officials are vague on the details “to avoid jeopardising the trial”.

On the face of it, then, IGG seems like a force for good. Yet dive deeper and a number of ethical, privacy and technological concerns appear.

In the case of the Golden State Killer, the handling of De Angelo’s DNA is said to have violated the terms and conditions of GEDmatch. At the time, all data was automatically opted-in to the public database but people could only upload DNA if they were authorised to do so by the owner. This didn’t apply in De Angelo’s case, which ethical campaigners, and his defence team, have leaned on since. GEDmatch subsequently changed its conditions so people now have to opt-in for their data to be open to officials.

The police can still serve court orders for online DNA data but, as a 23andMe spokesperson told E&T, “23andMe will exercise any available legal measures to object to a law enforcement request”. From its latest transparency report, 23andMe has only received 11 requests from US police (and zero in the UK) and the site has refused them all. This policy is the same for all major ancestry sites.

The situation in the UK is less murky. Ancestry sites are private and don’t allow DNA to be uploaded without the owner’s explicit consent, which they can withdraw at any time thanks to protections offered by the Data Protection Act 2018, and the EU General Data Protection Regulation (GDPR). UK and EU residents are also protected by Article 8 of the European Convention of Human Rights (ECHR) and the Human Rights Act 1998, which give members an automatic right to privacy. This is very different from the situation in the US.

“We have this unique posture in the States,” Andrew Crawford, senior policy counsel, Privacy and Data Project at the US Center for Democracy and Technology told E&T. “We don’t have comprehensive privacy laws at a federal level. We have sector-specific laws, and while they do a decent job of protecting people, big gaps exist.” For instance, the Health Insurance Portability and Accountability Act (HIPAA) only covers data held by certain entities like your doctor or health insurer. As the data from at-home DNA kits is held by consumer companies, HIPAA’s protections don’t apply by default.

The elephant in the room when discussing such protections is that they only apply to the person who owns the health data. It wouldn’t and couldn’t explicitly cover the thousands of relatives whose DNA is inadvertently uploaded, often without any knowledge, purely by genetic association. To put the scale of this into perspective, the average Brit is said to have 193,000 cousins.  

As Professor Teneille Brown from the University of Utah and Centre for Health Ethics, Arts and Humanities explains: “Even if the distant relative were told that her profile linked her to a suspected serial killer, it is hard to see how this revealed private information. It is potentially embarrassing and sensitive. But is it private, just because it was unknown?”

Such protections also assume that the user fully understands what they’re agreeing to regarding their own data, let alone other people’s. “The burden falls overwhelmingly on each consumer to do the legwork,” continues Crawford. “Most folks don’t have the time and stamina to digest each privacy policy. As a result people often decide to use a product or app without fully appreciating all the potential uses of their data.”

It’s obvious why criminals would object to their relatives uploading DNA to a database, but the pushback comes from the consequences it can also have on innocent people. DNA sites are being used to find biological parents of adoptees, or sperm and egg donors. This can be a positive use-case, but not every donor or parent will want to be found. The anonymity documents they signed effectively no longer apply.

The accuracy of the results has also been called into question. In 2014, officials using a public Mormon DNA database found what they thought was a match to DNA from a 1996 homicide. This led to a filmmaker called Michael Usry Jr being named a suspect. When a full DNA comparison was made, however, Usry was not a match and he was cleared of any involvement. He has been vocal in his criticism about the damage this false accusation caused, and continues to cause, since.

Elsewhere, in a 2018 study researchers found an “alarmingly high” number of false positives from at-home kits. Researchers ran full clinical diagnostics on people who had been told, by kits, they had genetic variants linked to chronic and fatal conditions. Almost half of these warnings (40 per cent) were false positives, meaning they had no such risk. A similar study, a year later, put the false discovery rate as high as 85 per cent in the case of very rare variants.

Yet as Professor Brown says: “The problem here is not unique to IGG. False arrests and false leads are inevitable in any investigation. Far from trampling on privacy rights, IGG may actually reduce invasions of privacy and biased prosecutions by law enforcement.”

Such false accusations and false positives stem from the limitations of at-home DNA kits. The majority of these use autosomal DNA testing, a technique that analyses the DNA inherited from both parents to glean information about a person’s genetics. Due to the vast amount of mixed information, pinpointing specific details can be difficult. For instance, when a DNA test tells someone they’re 35 per cent Irish, what they’re saying is that of the DNA analysed, 35 per cent of it was most similar to that of an Irish person. It’s a best guess.

The chance of an autosomal DNA test accurately detecting a relative also decreases with the distance of the relationship. Most autosomal DNA ancestry tests predict an accuracy rate of 90-98 per cent when matching third cousins, but this drops to a 45-50 per cent chance of detecting a match with a fourth cousin and beyond.

That’s not to say other DNA tests aren’t free from issues. Richard III’s nephew was identified using mtDNA testing. This technology looks for shared mitochondrial DNA between people to trace their matrilineal lineage. Experts claim it wasn’t accurate to say the remains belonged to the king because “any male sharing a maternal ancestress in the direct female line could qualify”.

Y-DNA testing is another alternative. It’s popular among genealogical researchers because it traces paternal lineage. The Y-chromosome remains virtually unchanged as it’s passed from father to son, so Y-DNA testing makes it possible to go back as many as 25 generations almost in a straight line. Yet these tests only reveal male ancestors.

Advocates play down all these concerns by arguing that IGG is just one tool in the wider toolkit available to detectives. At the Golden State Killer conference, Sheriff John McGinnis noted that “although it was DNA that led us down the right road, there were a lot of places that road could have led”.

Barbara Rae-Venter, the woman who ultimately identified De Angelo using his DNA, writes in her book, ‘I Know Who You Are’ that, even after finding DNA matches, she still had to spend days building family trees until she could name two potential suspects – Joseph De Angelo and his brother John. Investigators then still had to cross-reference victim descriptions, scope out the brothers’ respective backgrounds, alibis and more.

“Genetic genealogy provides clues, but you would always want to use traditional genealogy documents when trying to prove those relationships,” Graham Holton, genealogy tutor at the University of Strathclyde, told E&T. “Documented sources are, and have been, the main source for genealogists for decades but there are deficiencies – records are incomplete, names are missing or the father listed on a birth certificate, as one example, may not be the biological father.” In this way, neither method is perfect, but when used together, they can complement each other.

Then there’s the argument that the good that comes from IGG outweighs the bad. As the BFEG adds in its study, the ethical issues surrounding the Golden State Killer’s DNA were ignored “because the ends justified the means”. That’s an opinion shared by GEDmatch’s founder, Rogers: “If [the critics of IGG] could see the emails from families that have had some closure, I can’t imagine anyone would say it’s the wrong thing to do.”

Similarly, Christi Guerrini, assistant professor at the Center for Medical Ethics and Health Policy at Baylor College of Medicine, told E&T that “by catching a perpetrator earlier in their career, it can help catch violent criminals before they become serial offenders”. This could cut crime rates, costs, and the time it takes for victims to get justice, while also reducing the number of innocent people who come under scrutiny. In this way, Guerrini posits IGG could even end up protecting privacy, not eroding it.

Underpinning all the conversations around IGG’s pros and cons, Brown believes something much larger is at play – our human tendency “to feel discomfort with not having control over who (or what) ‘speaks’ for us”.

“We have allowed far too much fear of genetic informants to take hold,” she says. In trying to ban IGG, privacy advocates have misdiagnosed the problems. “Rather than feeding into this unwarranted fear, we must directly address the risks at their roots – by shoring up consent and passing legislation aimed at prohibiting certain secondary uses of genetic data.”