The recent arrest of a serial killer was clever detective work, but the police method raises uncomfortable and urgent questions.

In 1976, a killer was stalking the streets of Sacramento County, California. Over the next decade he would terrorise the community, murder 12 people and rape nearly 50 women and girls in a string of crimes that went unsolved for more than 40 years.

Then, in April this year, police uploaded DNA data taken from the crime scene to a genealogy website called GEDmatch.

They got a hit.

The DNA matched relatives of 72-year-old Joseph James DeAngelo, an ex-cop, who has now been charged as the suspected “Golden State killer”. While roundly welcomed, the result had a mixed reception from genealogy enthusiasts.

“I’ll volunteer to give my DNA and out any of my cousins who may be rapist/murderers,” wrote one on a Facebook page, quoted by the New York Times.

But a troubled GEDmatch client said, “My relatives consented for their data to be used for genealogy but not for criminal investigations.”

Those concerns feature prominently in a new commentary, published in the journal Annals of Internal Medicine. Written by a team led by bioethicist Benjamin Berkman from the US National Institutes of Health, it raises deep questions about the ethics of using genealogy DNA data to solve crime.

“Despite the popularity of online genealogy services, it is unclear whether users of these sites understand that their genetic data are available to criminal investigators,” the authors write.

Read the full article in Cosmos magazine here