Op-Ed: Responsible AI Is Making American Communities Safer

Dallas is booming. But growth brings challenges along with opportunity, including new demands on the men and women responsible for keeping a major American city safe.

 

The good news is Dallas has made meaningful progress against violent crime and several major categories. But those gains should not obscure the problems that remain. Burglary of motor vehicles is up this year, including a particularly sharp increase in the Central Division, while shoplifting has also risen. Dallas has confronted serious fiscal pressures at the same time its police officers are being asked to do more.

 

That combination should sound familiar to communities across America. Too many crimes, too much information and too few hours for investigators to connect every potentially important piece of evidence.

 

There is an opportunity to change that equation. Artificial intelligence, deployed responsibly, can give law enforcement a powerful new tool to solve crimes and make communities safer.

 

Like DNA analysis and fingerprinting before it, AI can transform how police identify leads and connect information. But it must be governed by commonsense guardrails and human judgment. AI should never replace an investigator, determine guilt or become an excuse to weaken Americans' civil liberties. Used properly, however, it can help detectives do something fundamental to police work – connect the dots. 

 

You cannot solve crimes without connecting the dots, and you cannot connect them until the evidence has been collected and organized. That sounds simple. In practice, it is enormously difficult.

 

Investigators can work a case skillfully and still miss the person three blocks away – or three states away – who possesses an important piece of information. The clue needed to solve a Dallas homicide may sit in another Texas community or another state. A witness may not realize that a place, person or event from years ago has any connection to an unsolved crime.

 

Meanwhile, tip lines fill with noise. Cold-case units confront enormous quantities of information. Digital evidence keeps multiplying. The person with the answer often does not even know he or she has it.

 

The national numbers show why we need a better approach. Analysis of FBI Uniform Crime Report data indicates that more than 350,000 homicides went unsolved between 1965 and 2024. Homicide clearance rates have also declined substantially over the past five decades. We are solving a smaller share of murders than we once did, while unresolved cases continue accumulating.

 

Missing-person cases present a related challenge. At the end of 2025, FBI figures showed nearly 90,000 active missing-person records across the country. These are not simply entries in a database. Each represents a person and, in many cases, a family waiting for an answer.

 

Crime also imposes an enormous economic burden. Research by economist David A. Anderson has estimated the annual cost of crime in the United States in the trillions of dollars when costs to victims are included. Those costs appear in medical bills, lost wages, higher prices, reduced investment and neighborhoods forced to live with the consequences of violence and disorder.

 

Dallas illustrates both the challenge and the opportunity.

 

The Dallas-Fort Worth economy is one of America's great growth stories. But prosperity does not eliminate crime, nor does economic growth automatically give detectives unlimited time or resources. Even as several Dallas crime categories have fallen, other offenses have moved in the wrong direction. Vehicle break-ins, for example, increased citywide this year and rose considerably faster in the Central Division. 

 

At the same time, Dallas has faced significant fiscal pressure. The city confronted a multimillion-dollar budget shortfall in 2026, with higher-than-expected police and fire overtime among the contributing expenses. Dallas is now investing more in public safety and plans to add officers, but recruiting and training those officers takes time. Technology can help close the capacity gap today.

 

That is where responsible AI can make a difference.

 

Consider a model that already exists in another area of public safety. The Dru Sjodin National Sex Offender Public Website connects information across jurisdictions, allowing people to search a shared problem rather than guess which government silo contains the relevant record.

 

Unsolved crimes need the principle of connection. Law enforcement needs tools that can securely organize enormous quantities of legitimate case information, identify relationships that human investigators may not have the time or ability to see, and present those potential connections to trained professionals for evaluation.

 

That is the gap the National Center for Open & Unsolved Cases was created to help close.

 

Our AI-powered platform, HOLMES – developed with U.S. Department of Justice COPS Office grant support and sustained with private donations – is designed around that principle. Law enforcement can enter public case information and locations connected to cases. Citizens can enter places that have shaped their lives without needing to know which investigation, if any, those places might touch.

 

HOLMES looks for meaningful overlap and can surface potential leads through an anonymous tip process. The base platform remains free to law enforcement agencies and the public.

 

Above it sits AIM – AI Investigative Methodology – a framework of standards, training and tools intended to help detectives and analysts use artificial intelligence as a professional and defensible investigative practice.

 

The distinction matters.

 

Responsible AI in law enforcement should mean professional tools operating under clear rules, with trained investigators making the decisions. It should not mean shadow software on a personal phone, secret algorithms determining guilt or technology operating without accountability.

 

AI's greatest public-safety promise is not replacing police officers. It is multiplying what good police officers can accomplish. A detective still determines whether a lead makes sense and prosecutors still have to prove their cases, but AI can help those professionals search information at a scale and speed that no human being can match.

 

For a growing city like Dallas, that matters. And for departments across America facing aging cold cases, enormous volumes of digital evidence and finite investigative resources, it matters even more.

 

For decades, law enforcement's problem has been that the critical dot may exist somewhere without investigators knowing where to look. Responsible AI gives us an opportunity to finally connect it. For families waiting years for an answer, that connection can mean everything.