AI & the Fire Department

Consider yourself sitting in a lecture on lithium-ion battery fires at a fire chief workshop. You listen to the issues and challenges that the fire service faces. You recently began to play with artificial intelligence (AI) apps, so you decide to see what AI has to offer to this conversation. You ask ChatGPT, for example, to generate a standard operating procedure/standard operating guideline (SOP/SOG) for the fire service that’s based on NFPA standards. Within 90 seconds, you’re presented with a three-page SOP/SOG on the subject. Although it appears to be comprehensive, can it be trusted?

Members of the fire service must educate themselves on the effect that AI likely is to have on their organization and must help to establish the parameters for managing the use of this tool.

What is AI?

Simply put, AI is the capability for software to perform tasks that have been completed within the human work space. Although there are multiple types, two of the most frequently referenced are generative AI, which has the capability to generate audio, video and written content, and machine learning, which uses algorithms to detect patterns to make assumptions and/or predictions that are based on large data sets. For the purposes of this article, these formats are covered here.

Fire chiefs must begin to place expectations and guardrails around the use of AI in their organization to prevent negative downstream consequences. ChatGPT alone found its way to 100 million users in just two months and is likely to add $4.4 trillion to the global economy on an annual basis, according to the management consulting firm McKinsey & Company.

Fire service opportunities

Several areas of the delivery of emergency services likely will be affected positively by the use of AI. They include:

• Recruitment and hiring. The use of AI can assess impact hiring by assessing large data sets and résumés to match potential applicants and provide generated content for recruitment purposes.
• Training and continuing education. Generating content through AI software allows recruit and continuing education instructors to generate course content, syllabi and grading matrices that are aligned with the learning needs of students.
• Occupational wellness. AI software assists in assessing large data sets to determine trends in the occupational wellness of firefighters and EMS providers. The fire administration can use these data to complete risk assessment and fulfill future occupational wellness needs for their organization.
• Vehicle data for apparatus maintenance. AI software is capable of tracking vehicle conditions to help to identify trouble before a breakdown occurs.
• Policy development. As in the scenario that’s referenced above, AI can search multiple references to put together a policy or SOP/SOG that potentially eliminates hours of research. That said, this doesn’t reduce the need to review and verify the information that’s used to develop the policy.
• EMS/public health. As the complexity of out-of-hospital care changes and community risk reduction evolves, AI will use large amounts of data to determine where community needs are and the best strategies for implementation for harm-reduction. It will provide a street-level provider with the opportunity to quickly search for information that’s related to the understanding of complex disease processes or medical management. It also will be capable of providing data on a wide range of response information, such as toxicology and behavioral medicine. AI can be used to predict patterns of illness in communities, which can assist public health with early detection and treatment as a disease advances throughout communities.
• Customer service. As cities and fire chiefs look for additional ways to enhance the customer experience, AI already is used, but it has plenty of room for additional growth. This includes interactive chats when community members are on a department website, to script responses to frequently asked questions, which will allow human intelligence to be available for more complex public interactions.
• Emergency management. AI software can help to plan exercises around different topics for which emergency management professionals are asked to prepare. As well, AI can assist with geographic information and weather analysis to predict fire patterns, evacuation distances, and other response planning and logistic needs for communities that are under duress.
• Fire prevention. AI can be used to assess evacuation pathways and occupancy loads for buildings of varied construction and to assist in understanding how weather and fire load can affect structures during the fire prevention planning and inspection phase.
• Emergency dispatch and response. AI software is assisting the next-generation 9-1-1 conversation by diverting some unnecessary calls that are made to 9-1-1 to other appropriate agencies, by enhancing call triage, and by identifying emergency response plans that are based on current traffic, weather and road closures.

Fire service challenges

A fire chief faces challenges with the use of AI, but they are manageable.

Depending on what and how you ask AI to generate content, you might get different responses. This puts the onus on department personnel to verify the accuracy of the content. In addition, AI can generate meeting summary documents that create diagrams of common themes that are outlined throughout the session, including specific quotes from participants but without providing context to the quote or comment, potentially promoting misunderstanding.

Professional standards units and their investigations will be challenged by complaints that result from graphic-processing AI software that creates and replicates seemingly legitimate fake social media videos, etc. This might undermine the credibility of the organization or its personnel.

Content that’s generated from AI sources potentially is open to plagiarism allegations of student/instructor misconduct. This could be an issue for fire chiefs, particularly in departments that have their own fire/EMS training programs and/or have affiliation agreements with local colleges/universities.

The New York Times is in litigation with Microsoft, claiming that OpenAI and Microsoft violated copyright laws by focusing software to scrape information of the newspaper’s content without permission. Although Microsoft denies the allegations, it’s an example of which fire chiefs must be mindful as they navigate the use of the content that’s produced from AI sources and how it affects the fair use doctrine that involves copyrighted material.

Depending on the type and sponsor of software applications, information that’s generated could be specific to that platform and create a potential bias in the information that’s created.

Although considerable discussion has occurred about the benefits of using AI sources, it shouldn’t be discounted that bad actors can use the same content to create disruption, study-response planning for emergency services, evacuation routes, etc.

Because of the current lack of oversight and regulation, accountability for privacy is of concern. Data that are generated will persist well beyond those that created the content. Data can be repurposed for additional uses that weren’t intended. Data potentially could be collected in ways that target people or organizations.

To augment

As noted by McKinsey, generative AI can automate, augment and accelerate work without necessarily replacing the human capital. For a fire chief, AI can stimulate thoughts, ideas and conversation. The use of AI can promote more opportunity for thought leadership within organizations. AI should be promoted to foster deeper and more meaningful engagement of leadership, rather than just defer to it to complete a project to meet a deadline. Put another way, leadership will discuss and assess content that’s generated by AI, rather than just conclude that what’s generated is suitable for the future of their organization.

Fire chiefs must ensure that they use this technology wisely and in a way that ensures the accuracy of the products that are developed from this information. Failure to do so will result in a lack of trust within team members, organizations and communities.

This technology is here to stay and advancing quickly. As fire service leaders, it can’t be ignored as a fad.