Artificial intelligence (AI) is the most disruptive innovation in a generation. It is quickly becoming an essential component in many industries, including public safety.

However, these are still the nascent stages of AI adoption, and with that, come challenges. One is the so-called “black box,” problem, where human operators overseeing a system do not fully understand why the algorithms recommend a particular action. Data goes in one side and results come out the other, but it is not always clear what happens in the interim. Maybe the software perceived a pattern that humans could not detect, or perhaps the machine simply made a mistake during its calculations.

The ability to explain decision-making consistently is crucial for AI models. Getting this much-needed clarity will help organizations gain the context needed for risk management, and this is especially true for a sector like public safety on which lives depend. The public’s faith in first responders’ decision-making must be absolute.

AI solutions in public safety need to firmly keep the human element in the driver’s seat so that the decision reached is fully vetted and understood. Indeed, public safety agencies have an ethical imperative to do so. That doesn’t negate the role of AI, however, as it can be used to expedite information-gathering before a decision has been reached. This form of AI in public safety is called “assistive” AI; like the name suggests, assistive AI does not execute commands independently. Instead, it helps first responders do their jobs more quickly and efficiently as they receive and process information day in and day out.

A perfect example of this combination of technology-assistance and human decision-making is in Public Safety Answering Points (PSAPs), where call-takers serve as the public’s first point of contact in an emergency. PSAP call-takers and dispatchers must separate the signal from the noise, which is all the more difficult when a flood of information from frantic callers washes over them. When the clock is ticking, highlighting critical data points even a few seconds earlier can save lives, and assistive AI excels at such essential work.

Rather than replacing PSAP personnel’s intuition, assistive AI can augment it by detecting similarities, keywords, and relevant insights across requests for service. For instance, operators fielding dozens of calls during a major traffic accident might not realize that the truck involved was a “tanker truck,” a detail only shared with one of the call-takers on duty. This slight nuance could indicate that the vehicle was carrying potentially explosive or otherwise hazardous liquids. That information would be vital for the first responders arriving on-scene, especially if a fire started because of the incident. Assistive AI would immediately highlight this critical detail to PSAP staff, who could inform arriving first responders to take precautions.

Besides PSAPs, assistive AI can perform tasks like flagging gunshot detection to nearby police or coordinating multi-agency response efforts. By continually analyzing information faster than humans could alone, the technology spots anomalies and patterns, helping first responders see the unseen.

There are also other intriguing areas on the horizon where assistive AI could help, such as acting as a second set of eyes to detect when a first responder’s stress levels become unmanageable. Public safety professionals routinely work 12-hour shifts while viewing traumatic scenes that adversely affect their mental health. Because of this, call-takers face increased fatigue, with a greater prevalence of post-traumatic stress disorder (PTSD) than other professions. Assistive AI could one day alert supervisors that a call-taker needs a break, allowing them to intervene while preserving employees’ energy and insight. By utilizing these helpful and life-saving resources, emergency communications centers could provide a more supportive environment that ultimately would make staff more effective.

Human intuition and decision-making will be the lynchpin of public safety for the foreseeable future because computers and software cannot wholly replace such complex judgements. But what AI can do is amplify these elements, better safeguarding the public, as well as alleviating the significant stress first responders grapple with day in and day out. In an industry where every second counts, AI that helps public safety officials do their job more efficiently, here and now, is not only welcome, but necessary.