Skip to main content

After the Golden State Killer's Apprehension, People Are Concerned About the Privacy of Genetic Tests. They Shouldn't Be.

The Joseph DeAngelo case has raised concerns about the privacy of data produced by consumer genetic tests. But really, law enforcement should be using DNA evidence more often.
Joseph DeAngelo appears in court for his arraignment on April 27th, 2018, in Sacramento, California.

Joseph DeAngelo appears in court for his arraignment on April 27th, 2018, in Sacramento, California.

Last month, police in Sacramento arrested a man for a series of rapes and murders that had gone unsolved for four decades. The break in the case came when investigators searched a public DNA database and found a partial match to DNA found at the crime scenes. That clue ultimately led investigators to Joseph DeAngelo, a former police officer who stands accused of being the notorious Golden State Killer, responsible for nearly 50 rapes and 12 murders spanning the 1970s and '80s.*

This seems like a clear-cut case of good news—a violent criminal brought to justice thanks to DNA evidence. But the Golden State Killer case has raised concerns about the privacy of data produced by consumer genetic tests, like those of 23andMe and "People who submit DNA for ancestors testing are unwittingly becoming genetic informants on their innocent family," a Maryland public defense attorney told USA Today. He worries that broad DNA searches for relatives of criminal suspects could mistakenly send police after the wrong person. Vice's Jason Koebler wrote, "[I]t should frighten you that police used an open-source genetic database to do it." Koebler argues that people should be wary of letting their genetic information end up in any large database—open source or not—that could be accessed by police and potentially misused.

If you had 23andMe or analyze your DNA, how worried should you be about your data? Or if you've thought about taking one of these genetic tests, should you reconsider?

It's an uncomfortable fact that when we use modern technology like cell phones and Facebook, we're constantly leaving a detailed record of ourselves—a record that police can use to learn about some of the most private aspects of our lives, often without a warrant. By having our DNA analyzed and stored someplace where it could be searched or subpoenaed, aren't we giving the police yet one more opportunity to invade our privacy?

But before you decide to delete your account, take a step back and consider how, exactly, your privacy would be violated if the police started routinely searching public genetic databases for matches to crime scene DNA. One concern—common but wrong—is that investigators might come up with a false match between the crime scene DNA and the DNA of an innocent person. This kind of false match is, in fact, essentially impossible, given the type of high-resolution data that is in a genetics ancestry database. It's true that consumer genetic tests aren't perfectly accurate in some ways—they don't always get your ancestry right, and they can be wrong about your risk for heart disease or cancer. However, estimating your genetic ancestry or predicting your cancer risk is a much harder problem than distinguishing one person from another. If they have a good sample, police aren't going to falsely match a DNA profile from a crime scene with a completely unrelated person in an ancestry database.

A more likely scenario is what happened in the Golden State Killer case: Investigators find a partial match, meaning that someone in the DNA database is unambiguously related to whoever left their DNA at the crime scene. In this way, innocent relatives of a suspect—possibly a mistaken suspect—could be drawn into an investigation simply by virtue of their genetic relationship. This does raise ethical concerns, including the question of whether pursuing such familial DNA matches is a violation of the Fourth Amendment's ban on unreasonable searches. But even in the absence of DNA evidence, relatives of criminal suspects sometimes are questioned by police as they search for leads. It's not clear why a genetic database search for familial matches is fundamentally different, ethically or constitutionally.

The biggest risk of letting the police routinely search genetic ancestry sites is that DNA found at a crime scene isn't always the DNA of the perpetrator or the victim. This is how DNA evidence in criminal cases is most likely to implicate innocent people. Investigators, putting too much confidence in the DNA evidence in hand, sometimes aggressively seek out further circumstantial evidence to pin a suspect to the crime, while ignoring exculpatory evidence. This is not a trivial risk. We leave traces of our DNA almost everywhere we've been. In some cases, one person's DNA can be transferred from one object to another by someone else. A chaotic crime scene could show DNA traces of many individuals—even on the murder weapon. When investigators wrongly assume they have the perpetrator's DNA, an innocent person can be convicted on what seems to be very compelling physical evidence.

This is what happened to American college student Amanda Knox, who was wrongly accused and convicted of murdering her housemate during a semester abroad in Italy. More commonly, the innocent person is someone more vulnerable—like Lukis Anderson, a homeless resident of San Jose who had a criminal record. Anderson's DNA ended up near a murder victim, most likely transferred there by paramedics who responded to the crime scene soon after taking Anderson to the hospital in an unrelated incident. Investigators thought Anderson was their man, until hospital records proved Anderson's alibi.

It's frightening when the wrong person is accused of a crime, but let's be real: The largely college-educated, middle-class customers of sites like and 23andMe are unlikely to be innocently drawn into a murder investigation just because their DNA is in a database that the police might search. When police collect DNA at a crime scene, the first place they look for a match is the Federal Bureau of Investigation's National DNA Index, which contains DNA profiles of over 16 million people. The FBI has this data because all 50 states collect and store DNA data of convicted felons, and most states allow investigators to collect DNA from people at the time of arrest for some types of crimes. The NDI is a powerful crime-fighting tool: The FBI reports that, as of March of 2018, this database has assisted in nearly 400,000 investigations.

But the NDI also represents one of the biggest problems with the way DNA is used in criminal justice. The NDI is the primary database searched for matches to crime scene DNA, which means that police first scan through a list of people with prior criminal records. This has some merit—repeat offenders are more likely to be responsible for a new crime than someone with no prior arrests. But, given America's racially biased mass incarceration problem, using the NDI means that the burden of misused DNA evidence or intrusive familial searches is more likely to fall on vulnerable people, often with criminal records and from minority populations—people who are not in a good position to defend themselves from false charges. If police used DNA databases representing a broader cross-section of Americans, they might solve more crimes and go after innocent people less often.

Another issue with DNA forensics: The technology used by forensic labs is more than a decade out of date with modern biotechnology. Forensic labs do not use the techniques and instruments that are now standard in clinical genetics; instead, these labs rely on an older technology that generates lower-information DNA profiles. That older technology works just fine when you have a good DNA sample, and it's also fast and cheap—important for underfunded crime labs that need to return results quickly. But that older technology is less accurate when used on very small or low-quality samples. Newer DNA technologies, capable of producing good data from difficult samples like 50,000-year-old Neanderthal bones or fetal DNA fragments circulating in a mother's blood, could improve the capacities of forensic genetics. Unfortunately, while the federal government invests billions of dollars in medical genetics, it spends just a fraction of that on research in forensic science. Under pressure to get results quickly and cheaply, and also facing a large backlog of DNA samples, forensic labs have little choice but to stick with the technology they've been using for decades.

This is a missed opportunity, because, with better DNA forensics, not only would more crimes would be solved, but also fewer innocent people would be convicted. The Cardozo School of Law's Innocence Project, which works to identify and free wrongly convicted people, reports that DNA evidence has exonerated 356 people serving prison time for a crime they didn't commit. Moreover, tens of thousands of suspects have been cleared after DNA evidence showed they were innocent.

Rather than worry about whether police will look at our 23andMe data, we should be concerned that DNA evidence isn't used as effectively as it could be to put violent criminals behind bars and keep innocent people out of jail.

*Update—May 16th, 2018: This post has been updated to reflect the fact that Joseph DeAngelo was fired from the Auburn Police Department after he was accused of shoplifting. He did not retire from the department, as we originally reported.