Murder & Artificial Intelligence AI

Murder and AI

Today’s artificial intelligence (AI) technology does offer criminal investigators greater insight into how to decrease serial killings and murders. First off, AI can be used to identify patterns in killings and to scan and process data more quickly than any human investigator can do (Garza & Martinez, 2019). So long as the algorithms are programmed to look for potential clues or patterns, AI can be of great utility in preventing crimes and reducing the spate of killings historically associated with serial killers. Secondly, data scientists can and machine learning can help decision makers know where to focus resources so as to catch a killer, as Brown (2017) points out. AI can help police sort through video of crimes, analyze a crime scene, compare crime scenes, and predict future crime spots (Sabnis & Khurana, 2018). Had AI technology been around at the time of Ted Bundy or the Atlanta Child Murders probably would not have happened for as many years as they did occur.

As Brown (2017) explains, data analytics is really at the heart of what policing is really all about when it comes to crime prevention. The data is what shows patterns, and it is what allows for predictions to be made about where to concentrate police forces. The data shows that certain crimes are more likely to occur in certain conditions, i.e., at night, in particular neighborhoods, and so on. If police are looking for a suspect or are looking to stop a type of crime, like serial killing, they can enter the data they have into an AI program and see what predictions can be made, what trends appear, and what patterns become obvious. Data analysts already do this type of work, but machine learning and AI tools can help analysts do it even more efficiently. Humans work slow, but machines—once programmed—can work very fast. And when time is of the essence, as it always in a murder case, the more machines one has working on a case—such as the Bundy case or the ATKID murders—the more likely it is that the next murder may be prevented.

AI helps with intelligence work because it does a lot of processing of information that humans cannot do on their own. AI technology can be “fed” data from prior serial killing sprees, so that it “understands” the defining characteristics and features of this type of crime. Based on that intelligence, it can create models that enable police to understand current crimes and how they might play out. Just like meteorologists used predictive modeling to make weather forecasts with good accuracy, investigators can use AI modeling to make forecasts about crime. AI can help police understand everything from motive to where a killer might try to strike next. This technology can facilitate investigations as well as police who are trying to prevent a future attack from occurring. As Garza and Martinez (2019) explain, “through the use of machine learning, we apply random forest algorithms to create predictive models to analyze more than a thousand serial killers, and be able to predict motivation, potential number of victims, and the sex of serial killers” (p. 1). Had this kind of technology been in existence in the 20th century, many killings probably could have been prevented.

One can actually see how AI can help with a real case: the Zodiac killer, for example, used ciphers to try to get police to figure out who he was at the time. As Sabnis and Khurana (2018) explain, “only one of the 4 ciphers has been solved. Now, a computer called CARMEL has been tuned to think like the Zodiac killer. Cipher decryption has the following steps; we first try to interpret the cipher and then use either appropriate keys or design a new one to fit the cipher. So CARMEL is going through the internet to get billions and billions of keys so as to try and interpret the Zodiac cipher.” No human or team of humans could possibly gather data the way that CARMEL is able to do; it is an AI tool that does the work of hundreds or even thousands of data analysts—all with the purpose of catching the Zodiac killer.

Yet AI technology has even more mundane applications—such as analyzing gunshots, interpreting grainy or blurry security footage, facial recognition, voice recognition, predicting future crime spots, and so on. The more that technology becomes a part of everyday life, the more that AI will play a part in stopping or preventing crimes—the reason being that technology is all around us and so too is AI. AI helps police to read between the lines, to see behind the curtain, to highlight what is not evident to human intelligence. Of course, for AI to work effectively it has to be programmed effectively, which means there is always going to be a human intelligence behind AI—but it is no different from the kind of human intelligence required to build a car: the human builds the car, but once it is built the human is not responsible for transporting the passenger; the car does the work. The same is true of AI: the human must put it together; but once it is put together, the AI technology does the work.