From Emergency Response to AI Security Engineering

Over 20+ years in the fire service, I’ve progressed from firefighter (truck and engine companies) to hazardous materials technician (15+ years), training division instructor (technical rescue and hazmat response), and now fire captain. I’ve spent my career responding to emergencies, leading teams under pressure, and making split-second decisions that protect lives and property. As my fire service career winds down, I’m preparing to embark on a new challenge that leverages this experience in an unexpected way: AI security engineering, with a focus on emergency services.

Why AI Security?

The fire service is rapidly adopting AI-powered systems—from dispatch optimization and predictive analytics to computer-aided dispatch (CAD) and electronic patient care reporting (ePCR) systems. But with this adoption comes risk: these systems handle sensitive patient data, personnel information, and tactical intelligence. A security breach or adversarial attack on these systems doesn’t just compromise data—it could impact emergency response times and patient care.

I’ve become fascinated by the intersection of AI and cybersecurity, particularly the unique challenges of securing AI systems in high-stakes operational environments. The questions around AI safety, adversarial robustness, and secure deployment remind me of the safety protocols and risk management fundamental to firefighting. Just as we plan for worst-case scenarios in emergency services, I believe we need the same rigor and foresight in developing and securing AI systems.

Why the Parallels Matter

The overlap between emergency services and AI security runs deeper than it might first appear:

Understanding the actual problems — I’ve experienced firsthand the pain points that plague emergency response: training systems that don’t reflect real-world complexity, operational workflows that create friction under pressure, and documentation requirements that pull focus from critical tasks. I also understand the systems we rely on—CAD platforms, ePCR software, incident reporting databases—and where vulnerabilities could have operational consequences. This isn’t theoretical knowledge from user interviews—it’s two decades of living with these systems daily.

Risk mitigation mindset — Fifteen years as a hazmat technician taught me to think in failure modes, cascading effects, and layers of protection. In AI security, these same principles apply: anticipating attack vectors, planning for adversarial scenarios, building defense in depth, and ensuring systems fail safely rather than catastrophically.

Teaching complex concepts — As a training division instructor for technical rescue and hazmat response, I’ve learned to break down intricate procedures into understandable frameworks. This skill translates directly to explaining security risks to non-technical stakeholders, documenting threat models, and bridging the gap between security teams and operational users.

Decision-making under uncertainty — Emergency scenes rarely offer complete information. You assess what you know, identify what matters most, and adapt as situations evolve. Cybersecurity—especially in defending AI systems—demands the same comfort with ambiguity, pattern recognition under pressure, and iterative problem-solving.

What I’m Doing

I’m currently pursuing a Master’s degree in Artificial Intelligence with a focus on AI security and cybersecurity. My coursework spans machine learning, neural networks, adversarial machine learning, secure AI deployment, and cybersecurity fundamentals. Beyond academics, I’m focused on understanding the specific security challenges facing fire and EMS agencies as they adopt AI technologies.

My goal is to help emergency services organizations adopt AI safely and effectively—ensuring these powerful tools enhance rather than compromise their missions. This blog serves as my learning journal, project portfolio, and a space to explore how emergency services experience informs my approach to AI security challenges.

What You’ll Find Here

• Project showcases: AI security projects and tools I’m building, particularly for emergency services applications
• Technical deep-dives: Breaking down AI security concepts, threat models, and defensive strategies
• Learning reflections: Honest accounts of successes, struggles, and breakthroughs in my transition
• Domain-specific insights: How fire service operational requirements intersect with AI security concerns

Let’s Connect

I’m eager to connect with others in the AI security community, fellow career changers, emergency services technology professionals, or anyone interested in the intersection of public safety and AI.

Reach out via GitHub or email at jacklexidrew@gmail.com—I’d love to connect.

---

“The best time to plant a tree was 20 years ago. The second best time is now.”