AI as an Investigative Partner: What a 33-Year Cold Case Shows Us

Some cases outlast the investigators who worked them. Witnesses age. Evidence degrades. And the file eventually gets boxed up and shelved. For decades, that was the end of the story.

Today, AI is changing what’s possible in investigative work. According to CentralSquare’s AI Insights Report, 80% of public sector leaders are extremely or very interested in adopting AI. Early adopters report roughly twice the results of agencies still waiting, while late adopters risk falling behind on investigative efficiency and even case closures.

Madison County Sheriff’s Office in Illinois showed what’s possible with AI for detectives. With the help of Centerline AI, investigators solved a 33-year-old cold case. This article tells that story, and what it means for agencies considering detective AI as a “partner.” Keep reading to see how it works.

The 1993 Cold Case That Took 33 Years to Solve

On November 9, 1993, Randy Gail Sperino—a 34-year-old mother—was found bludgeoned to death in a field in Granite City, Illinois. Her killer left behind DNA evidence, and investigators had a lead. What they didn’t have, for the next 33 years, was a match.

Starting in 1998, that DNA was run through CODIS repeatedly. Nothing came back. The case went cold, but it was never forgotten.

Madison County Sheriff, Jeff Connor, grew up two blocks from Sperino’s childhood home and worked for the Granite City Police Department the night she was killed. He had kept in touch with Sperino’s son, Wes, over the years. Although Connor didn’t have answers, he wanted to assure Wes the department hadn’t stopped looking.

About a year and a half ago, Captain Tim Lawrence’s investigations division reopened the case with new tools and a different approach.

How Centerline AI Became an Investigative Partner

The Sperino case file was massive. It included over 2,000 pages of interviews and reports, more than 300 pieces of physical evidence, and documents from typewriters that were still in use until 2008. Records clerks spent two weeks digitizing the file before Detective Jake Svoboda built the case inside Centerline AI.

Once the file was in Centerline AI, investigators worked through it by asking questions and following leads. They queried three decades of documentation conversationally, without manually reviewing every page.

“You just ask it questions like you’re in a text message almost, and you can spin yourself up really quick,” said Captain Lawrence.

Svoboda rebuilt the original investigators’ timeline, cross-referenced names across 30 years of interviews, and flagged contradictions between witness statements. AI flagged the discrepancies, and Svoboda determined which ones were significant and why.

“Every detective has their own personal analyst. It does not replace an investigator. It’s only as accurate as the information you give it.”

– Detective Jake Svoboda, Madison County Sheriff’s Office, Illinois

DNA, Forensic Genetic Genealogy, and the Break in the Case

Centerline AI gave investigators command of the file. Forensic genetic genealogy gave them the answer, buried in the DNA.

Traditional CODIS searches rely on a direct match against profiles of known offenders already in the system. Zigler had no prior record, so his DNA was never in the database.

Forensic genetic genealogy works differently. By sequencing crime scene DNA and comparing it against genealogy databases, investigators can identify distant relatives and build a family tree backward to a suspect, even one with no prior record. The technique wasn’t widely available until around 2018.

Investigators submitted the 1993 DNA evidence to Othram, a Texas-based forensic genealogy lab, and the analysis pointed to Albert “Buddy” Zigler, 70, of Caseyville, Illinois. When investigators approached him, he confessed.

On May 26, 2026, Zigler was charged with two counts of first-degree murder. The case is still moving through the criminal process.

For Wes Sperino, the arrest ended a 33-year wait for answers. “She was just a mom who loved me. I miss my mom,” he said after the arrest. “The crime is finally solved.”

What This Means for Law Enforcement Agencies Weighing AI

For agencies still on the fence about police AI, Madison County’s experience is worth examining closely.

Most agencies don’t have a dedicated cold case unit. According to a report in Police Chief Magazine, officers spend three or more hours per shift on documentation, leaving little time for unsolved cases.

At Madison County Sheriff’s Office, however, Captain Lawrence now assigns unsolved cases 20 to 60 years old without adding to anyone’s caseload. Reviewing 2,000 pages inside Centerline AI takes days instead of weeks, making it operationally viable to reopen them.

Legal uncertainty is one of the top barriers to public safety AI adoption. According to the AI Insights Report, 25% of respondents aren’t confident about how AI evidence will be treated in court.

Madison County’s Third Judicial Circuit has already issued two standing orders permitting its use. Centerline AI is CJIS-compliant and logs who viewed what and when. The state’s attorney’s office purchased its own license to keep detective notes and prosecutor trial prep separate for discovery purposes.

AI as an Investigative Partner for Cold Cases

Madison County didn’t have to hire a new detective to solve the case. They simply had to equip their officers with an AI investigative partner that could work through three decades of paperwork fast enough to still be useful. Their investigators did the rest.

The case is a reminder that justice has no statute of limitations. The only thing that changes is whether investigators have the tools to keep looking.

If your agency has cases sitting in storage that deserve a second look, watch the full webinar with Madison County’s investigators to see how an AI cold case investigation actually works. Or schedule a demo today to see how Centerline AI can streamline documentation and help your agency pursue justice more effectively.

Schedule a Demo Today