The tech helping crack decades-old cold cases

Randy Gail Sperino vanished on the evening of 9 November 1993 in Granite City, Illinois. Hours later, her body was found in a nearby field. For years, investigators had DNA from the scene, but the usual law enforcement databases turned up nothing. The case became one of more than 250,000 unsolved homicides in the US.

Last year, Madison County detectives tried something different. They sent the genetic evidence to Othram, a Houston company that uses DNA technology in cold and current cases. The method they leaned on was investigative genetic genealogy, the same broad approach used to help identify the Golden State Killer in 2018. Instead of looking for an exact match, it uses DNA from relatives to narrow down who an unknown person might be.

That lead pointed investigators toward Albert “Buddy” Zigler, now 70, who later confessed. He was charged in May this year. For Gail Sperino’s son, Wes, the result brought the obvious kind of relief: finally knowing who did it.

But the DNA work was only part of the story. The Madison County team also used Centerline AI, a platform from Florida software company CentralSquare, to sort through more than 2,000 pages of records from the case. It helped them build timelines, organise witness information and decide what evidence should be tested. Lawrence said the team could even interact with it like a chatbot, and later use it to think through whether Zigler was telling the truth.

Still, the captain was clear about the limits. AI can help detectives work faster, but it doesn’t replace them. Genetic genealogy has limits too: if there’s no DNA, there’s nothing to work with. Privacy is another issue, since one person’s consent to use a database doesn’t automatically cover their relatives. And then there’s money. Agencies want the tools, Mittelman says, but funding is thin, so some cases depend on donations or crowdfunding.