How AI is Giving Cold Cases a Second Chance

In 1993, Randy Gail Sperino was beaten to death and left in a rural field near Granite City, Illinois. Her killer vanished. For 33 years, investigators chased leads, re-interviewed witnesses, and ran DNA evidence through CODIS—with nothing to show for it.

In 2026, that changed.

A combination of forensic genealogy DNA and AI helped crack the case. Investigators uploaded more than 2,000 pages of case files into Centerline AI and began querying them conversationally. It allowed them to get up to speed on three decades of evidence in a fraction of the time it would have taken manually.

The case is a signal, not an outlier. AI for law enforcement investigations is fundamentally changing what’s possible in investigative work, including cases long thought unsolvable. Madison County Sheriff’s Office is proof of what’s possible.

Agencies that integrate AI into their records and case management infrastructure aren’t just improving active investigations. They’re reopening doors on justice that were once considered closed.

The Cold Case Problem Has Always Been a Data Problem

Some cold cases aren’t unsolvable. The answers are just buried. Information is siloed across thousands of pages and generations of officers, making it difficult to connect the right dots.

One agency’s experience illustrates the scope of the problem. A 2022 DOJ report found that the LAPD had scanned roughly 4,000 cold case binders, but manual data entry was still incomplete for nearly 700 of them as of early 2022. So the information inside wasn’t yet searchable or usable. That’s one of the most well-resourced agencies in the country, still manually processing files from decades ago.

Manually reviewing a cold case file can take an experienced investigator several hours. For someone less experienced, it can take up to a week. Legacy records management systems (RMS) were built for storage, not retrieval or analysis.

The administrative burden compounds the problem. According to a report in Police Chief Magazine, 56% of law enforcement professionals spend three or more hours per shift on documentation rather than actively solving crimes. For detectives juggling cold cases alongside active caseloads, every hour on paperwork is an hour away from the case.

The question isn’t whether the evidence exists. In many cases, it does. The question is whether investigators can surface it fast enough to act on it.

What the Madison County Case Demonstrates

The Sperino case offers a blueprint for how AI fits into a real investigation, one that keeps traditional methods at the center.

Investigators used interviews, DNA testing, and field work alongside Centerline AI, which let them query their entire case file as a single data source. Handwritten notes and recorded interviews—two formats that have historically required manual, time-intensive review—were suddenly searchable.

Captain Lawrence described using Centerline as intuitive. For a case spanning 33 years and more than 2,000 pages, that ease of use made all the difference.

“With Centerline…you put that data source (the case file) into the software, and you just pretty much ask it questions like you’re in a text message almost, and you can spin yourself up really quick.”

— Captain Tim Lawrence, Madison County Sheriff’s Office

Madison County investigators understand the purpose of AI. It doesn’t replace detectives; it surfaces connections and details buried in decades of documentation. Skilled investigators still have to verify and act on them.

Rob Sanderson, a former officer who now works with Centerline, said it plainly: “It’s definitely not a replacement for good police work. I want to call that out. The detectives in Madison County did a wonderful job and just happened to leverage this tool.”

What Modern Investigative Technology Should Deliver in 2026

The Madison County case illustrates what’s possible. But it also raises a practical question for every agency: does your current infrastructure support this kind of work?

Modern detective tools and case management software should deliver five things.

Unified Case Data

Case data should be accessible in one place. When incident reports, prior call history, interview notes, and records are spread across disconnected systems, investigators have to piece the puzzle together manually.

AI-Assisted Review

Detectives should be able to query large case files, identify patterns, and surface inconsistencies without manual document-by-document review. That’s the capability that made the Sperino case breakthrough possible.

Integrated Systems

CAD, RMS, and case management should function as one system, not three. Data entered into any of these tools should flow automatically to the rest. No duplicate entry. No version conflicts. No time wasted reconciling disparate systems.

Offline and Mobile Capability

Detectives work in the field. Their tools should too. Whether reviewing case files between interviews, accessing records at a scene, or running searches away from the office, investigators need reliable access to their systems without being tethered to a desk.

CJIS-Compliant Security

Cold case files contain decades of sensitive information: witness identities, victim details, confidential informant records, and more. CJIS compliance ensures this data is handled, stored, and accessed in ways that protect both the investigation and the people involved.

Centerline’s results are best described by the agencies using it. And while it’s proven itself as cold case software, its impact reaches well beyond unsolved cases.

“Tasks that previously consumed hours throughout my two decades of law enforcement are now completed in minutes, cutting 90% off the redaction process.”

— Captain Barry Morgan of the Mississippi County Sheriff’s Office.

The Recruitment and Retention Angle

Outdated investigative tools don’t just slow cases down. They may be driving candidates away too.

Investigative roles are already hard to fill. When the tools investigators rely on can’t keep up with their cases, it sends a message—this agency doesn’t invest in its people. Imagine spending an hour digging through a physical binder to find a witness statement that a simple query could surface in seconds.

Modern investigators grew up with smartphones and consumer AI tools. They know what good technology feels like. When their professional tools are a generation behind, they notice.

Agencies investing in investigative workflow software and public safety AI are sending the opposite signal. They’re telling recruits and veterans that they value their work and time enough to invest in the right tools.

In a competitive labor market, that message matters more than most agencies realize.

Questions to Ask Your Agency

The Madison County case is a useful lens for any agency evaluating police AI or investigative technology. Here are five questions worth asking before your next cold case review.

1. How long does it take a new detective to get fully oriented in an inherited cold case? If the answer is weeks, that time is pulled directly from active caseloads. And it suggests your records infrastructure wasn’t built for this kind of work.
2. Can investigators query case files, interviews, and records from a single interface? If they’re toggling between disconnected systems, it means they’re spending investigative time on logistics instead of leads.
3. If we reopened our five oldest unsolved cases tomorrow, what would the first 40 hours look like? The answer reveals a lot about where your agency’s investigative workflow stands today.
4. Does our current RMS make it easy to connect dots across cases, officers, and years? If that still requires manual effort, the answer to your next cold case may already be in your files, waiting to be found.
5. What would our agency gain if investigators recovered even 20% of their administrative time? It’s worth doing the math. By reducing administrative work, your agency could review more cases, and pursue more leads.

Built for Detective Work and Beyond

The Madison County case is a reminder that justice has no statute of limitations. The right tools can reopen cases that decades of traditional methods couldn’t close. That matters deeply to the families still waiting for answers.

AI doesn’t replace the detectives who do the work. It clears the path so they can do it faster, with more confidence, and without losing weeks to manual document review.

CentralSquare’s public safety software is fully integrated and built for how agencies actually operate. From the 911 call to CAD, RMS, and case management, data flows across the platform so investigators have everything they need in one place.

Ready to see what modern investigative technology looks like for your agency? In a 30-minute conversation, we can walk through your current workflow, your integration requirements, and your agency’s path to modernization. Schedule a demo today.