How Can AI Help with Real-Life Cold Case Files?AI doesn’t create new ideas in police work; rather, it does the work that police, who must move on to urgent, fresh cases, don’t have time to do
When no new leads emerge in a murder or missing persons investigation, police must shift their resources to cases that offer new information. Currently, the FBI Uniform Crime Report keeps an estimated 250,000 cold cases on file, on which it co-operates with local law enforcement. Recent developments in AI, however, have shed light on some of these old and cold cases.
For example, in 2018, DNA evidence submitted through a popular genealogy database enabled charges to be laid in more than a dozen cases involving the Golden State Killer who had been living “ under the radar” for decades. According to a recent article in Forensic Science International, in 2018, direct-to-consumer (DTC) genetic genealogy databases were used to solve over fifty cold cases of either murder or missing persons.
These uses of AI raise privacy and ethics issues, of course. But they raise those issues precisely because the method actually provides potential new information.
AI doesn’t create new ideas in police work; rather, it does the work that police, who must move on to urgent, fresh cases, don’t have time to do. For example, it can grade information in the files for unseen clues, global positioning, trace evidence, and even chemical analysis. Detectives and volunteers can then have a fresh look once the AI results are merged with DNA lab results, facial recognition, and human analysis to see if a new pattern is worth pursuing.
For example, there is the case of the West Mesa Bone Collector. Between 2001 and 2005, eleven women were buried in an undeveloped tract of land outside Albuquerque, New Mexico. A real estate developer had begun construction there a couple of years later but had to stop because of the mortgage crisis of 2008. In 2009, floods caused human remains to surface in the disturbed soil.
The crime is still unsolved but DNA technology, along with witnesses and satellite imagery, has provided some clues. For example, in 2010, two of the skeletal remains were determined to be misidentified because those two women were identified as alive. Some persons of interest have also been identified as possible suspects.
AI applications in the West Mesa case and others could be sifted and evaluated for false positives and misidentifications. In the past, old and cold cases have often been considered too complex to reopen. But AI might simplify the process and maintain the safe-keeping of relevant information, preventing loss or manipulation.
The globe is a big place, yet it has shrunk in certain ways, due to the ease of international travel and instant communication. INTERPOL, the world’s biggest repository of police cooperation and crime control, is pursuing strategies, along with the United Nations Interregional Crime and Justice Research Institute, in AI, robotics, future-casting, and facial recognition to keep pace.
They have their work cut out for them. The suspects in the West Mesa case may well have fled to a foreign country where extradition agreements with the United States are not recognized. In any event, there are countless unsolved cases that have international implications, such as the murder of British journalist Jill Dando in London (1999), the death of the Setagaya family in Tokyo (2001), or even the former Vladimir Putin aide, Mikhail Lesin (2015) found dead in a hotel room in DuPont Circle, Washington, DC.
AI could very well prove useful in the case of the Zodiac Killer (late 60s in San Francisco), Tupac Shakur (Las Vegas, 1996) and even JonBenét Ramsey (Boulder, 1996). Did Natalie Wood (California, 1981) really drown? Did D. B. Cooper (Washington State, 1971) really hijack a plane and jump out to never be found? AI may never solve the case of Jimmy Hoffa (Detroit area, 1975); however digitizing the evidence, testimonies, information about suspects could turn up some useful patterns.
Here’s what police are doing with AI today, and some reactions to it:
Here are some other fields in which AI can help solve longstanding puzzles:
Can AI help us decipher lost languages? Yes, sometimes. It depends mainly on the reasons we haven’t yet deciphered ancient texts.
Can AI prove that Shakespeare had ghostwriters? An author’s unique style is like a fingerprint. AI can fill it in and give us some patterns to look at.
Does AI challenge Biblical archeology? Many surviving documents are so damaged that they cannot be read using traditional methods. Recently, AI deciphered a nearly destroyed Biblical manuscript, showing that it was the earliest surviving copy.