WILL THAULT: From crime prevention to crime prediction

The year is 2054. Thanks to an experimental, but highly effective law enforcement tool, crime has been virtually eliminated in Washington, D.C., bringing the murder rate down to zero for the past six years. The system involves the use of three psychics, called “Pre-Cogs,” who can predict murders before they happen. Based on their premonitions, the newly-found Pre Crime Police Department uses the Pre-Cogs’ visual projections as evidence to identify, arrest and convict these pre-criminals prior to the commitment of their capital offense.

This was the plotline of the 2002 Tom Cruise movie “Minority Report,” loosely based on a book by the same name written by Philip K. Dick in 1956. A tale of science fiction that may soon become science fact, not due to Pre-Cogs, but Artificial Intelligence (AI).

Much has been said about the so-called “AI Boom” as the world first gained access to large language models and AI chatbots, like ChatGPT, just 20-years after the sci-fi movie’s release.

As with any technological or industrial innovation throughout human history, great advances for good as well as horrific applications for evil developed over time. But never has the expansion of a new technology accelerated to levels that will soon outpace our ability to harness or control it. Just as with the “Space Race” of the ‘60s and ‘70s, many countries are now competing with the U.S. for global dominance in an effort to harness the awesome powers of AI.

Why? The future seems to hold promise for untold advances in health care, medical research, robotics, transportation, scientific discoveries, agriculture, enhanced cybersecurity, defense intelligence and logistics, data mining and predictive behavior modeling.

It’s this last AI “advancement,” predictive behavior modeling, that raises immediate ethical and privacy concerns. Integration of this advanced technology into law enforcement strategies is already well under way in the form of data analysis of surveillance systems, social media, criminal records, and sensors that help in identifying patterns and trends conventional detective work might overlook. Things like predictive policing using AI algorithms to analyze historical crime data, suspect identification by matching facial recognition from surveillance footage or public cameras with biometric databases, along with real time Video Analytics, Social Media/Open-Source Intelligence (OSINT), and network mapping using Geographic Information Systems (GIS).

Near future AI advancements will begin to move from predictive modeling into predictive policing as explored in “Minority Report.” Aaron Cohen, a former Israeli Defense Forces special ops veteran, made national news recently with claims to be building “the first real-time threat prediction engine for law enforcement and schools” called GIDEON, an AI-powered system for scanning the open web, looking for signs of radicalization like “tactical planning” and “extremist indoctrination.” He believes GIDEON will be able to identify potential threats before they materialize. If so, will this, or something like it, soon justify the enforcement of pre-crime arrests?

Whether true or not, there’s a growing industry of private surveillance companies already selling detailed information about you and me to who knows who. Our personal digital footprints are routinely being monitored and harvested with little oversight or consent. AI will only make it that much easier to gather, integrate and analyze vast amounts of data in minutes.

Good intentions invariably lead to unintentional results – or simply bad intentions. When it comes to law enforcement, predictive tools such as these could result in saving lives by stopping crime before it happens. But what of the dangers that may come from accidental or purposeful abuse of this power?

First and foremost, when does widespread AI surveillance cross the line by infringing on an individual’s privacy rights? Does profiling by gathering personal data, facial recognition and social media on law-abiding individuals who haven’t committed any crime send a chilling message to those who live in a free society such as ours?

Also, what about “garbage in/garbage out?” How does AI generate an objective outcome if the historical database source it draws from happens to be culturally imbalanced, resulting in unintentional racial profiling or the unfair targeting of marginalized communities?

And, as always, where there’s opportunity for good, evil will attempt to corrupt the system motivated by personal or political gain. An exasperated James Bond once said to arch-villain Dr. No, “World domination. The same old dream.” Today, AI’s the dream. In our rush to control it, what’s to prevent it from first controlling us?

Transparency, safeguards and accountability must be created either through industry self-regulation or government control to contain these expanding omniscient powers. Without guardrails, AI will ultimately grow beyond our ability to control it. On that day, it will choose on its own whether to continue to be a benevolent helper to humanity or a tyrannical beast.

Will Thault is a retired businessman who has worked in media and retail marketing. He worked with Ted Turner’s “Super Station” TBS in its early stages.