Emerging technology has resulted in an exponential increase in the collection of data for law enforcement investigations and intelligence. This creates a challenge for federal, state and local law enforcement agencies to process through the data to find key information and ultimately make timely and informed decisions.

That’s why the promise of artificial intelligence (AI) is so powerful and necessary for today’s law enforcement. AI has the potential to make law enforcement information and data analysis – of all types and from multiple sources – faster, more efficient, more directed, and ultimately significantly lead to more effective decision making. Law enforcement investigative activity is increasingly happening at the petabyte level and increasing every day.

AI fundamentally represents the next chapter by enabling agencies to make critical strategic and tactical decisions in protecting our nation and preventing harm to the American people.

Adapting to Emerging Threats

For example, the execution of a lawfully obtained search warrant has become an increasingly complex activity with the emergence of the Internet of Things. Cell phones, smart devices, and other video and audio collection capabilities present challenges for law enforcement in identifying and evaluating the multitude of data that could be pertinent to an investigation.

With the right application of AI towards video analysis, the timeline to review collected data decreases significantly – from hours and days of traditional on-hand analysis down to minutes. Details that would be near impossible to detect by a human eye now can become visible and apparent.

With the power of AI, merging these findings with other collected evidence now becomes an achievable task for faster investigative results – leading to more accurate and expedient intelligence and information sharing across law enforcement agencies.

Addressing Operational Challenges

AI can also be applied to address the operational challenges that confront law enforcement agencies today. One of these is finite resources – the vast number of threats confronted by law enforcement means agencies must be discerning regarding the utilization of their personnel. An analyst spending valuable time sorting through what turns out to be non-important information results in less time spent on more relevant and critical intelligence.

Large language models and AI-driven tools allow analysts to quickly synthesize data, ask questions of it, and generate new, deeper and more precise questions to ask. Analysts can move more quickly than ever and save more of their most precious resource, time, especially acute during complex cross-agency investigations.

Aided by automation and AI, analysts can perform predictive analytics, forecasting, classification, correlation and can better leverage textual data with large language models, natural language processing, and generative AI.

Responding to Mass Offender Events

A critical incident or other large-scale criminal activity frequently results in vast amounts of data that is lawfully collected over a relatively short period of time, with a pressing need for real-time analysis to identify and mitigate any additional threats.

AI can greatly assist law enforcement personnel to find the right data and cross-reference it with historical information contained within current intelligence holdings. It can then be enriched as additional data is lawfully collected, allowing a more streamlined process to identify previously unknown relationships, identify bad actors and make smart and well-informed operational decisions towards mitigating current and future threats.

Swiftly incorporating AI for activities like predictive analytics, trend spotting, and relationship mapping represent huge opportunities. Additional experimentation with AI will also assist to guide law enforcement agencies to find the most applicable and achievable use cases throughout their organizations.

Prioritizing Ethical AI

Ethical considerations of AI in law enforcement are always paramount.
1. There must be a “human in the loop” throughout an investigative process to avoid an overdependency on the AI alone.
2. Understandable guidelines need to be put in place and practiced regarding how, when and where AI becomes integrated within law enforcement activity.
3. And very importantly, the ability to explain and defend the use of AI during a criminal proceeding is tantamount to its successful utilization.

Ralph Tursi is a retired Special Agent for the Federal Bureau of Investigation (FBI). During his time at the FBI, Ralph was the deputy assistant director of the agency’s human resources division; the chief of East Asia intelligence in the agency’s counterintelligence division; an FBI senior executive to the National Security Council on interagency policy development; and served as an Assistant Legal Attache’ at the U.S. Embassy in Canberra, Australia.