Security by Design: Enhancing Human Judgment in an AI-Enabled World
Every day, our brains are processing tens of thousands of sensory cues, whether we’re walking down the street or working in a high-stakes environment such as an airport checkpoint, cargo terminal, or border crossing.
This constant scanning for potential danger is automatic, rooted in evolution: when faced with ambiguity, we’re wired to prioritize speed over accuracy. The brain’s threat detection system triggers a physiological response, what we often call “gut instinct” before the rational part of the brain has time to fully assess the situation. It’s a system designed for survival, not optimization.
In high-stimulus environments like security operations, relying solely on physiological responses can lead to false alarms, missed threats, and disrupted workflows. In security environments where both throughput and precision are mission-critical, understanding how the human brain works under pressure is not just interesting—it’s essential.
To understand how human psychology affects real-world performance, we must first explore the mechanics of sustained attention and how quickly it can falter under pressure.
Attention Under Pressure
Whether it’s evaluating X-ray images, scrutinizing cargo manifests, or observing surveillance feeds, sustained attention is vulnerable to vigilance decrement. This phenomenon describes the drop in sustained attention over time, especially in monotonous tasks. The drop in performance typically occurs after just 20 minutes of continuous attention—which is why many regulatory bodies such as the European Civil Aviation Conference (ECAC) limit image evaluation to that duration.
The problem becomes even more complex in low-prevalence environments, where genuine threats are rare. In airport and cargo screening, for instance, 99.9 percent of items screened are harmless. Over time, this low hit rate can cause the human brain to unconsciously tune out, making it more likely that the rare genuine threat will be missed. Understanding and designing for the limits of human attention is essential to keeping security both effective and efficient.
The drop in performance typically occurs after just 20 minutes of continuous attention.
Security professionals are often expected to make rapid decisions under pressure, whether it's how to respond to a suspicious object, a nervous passenger, or a momentary alarm. These decision-making moments are often explained through the lens of dual-process theory, which suggests we rely on two distinct thinking systems. The first, System 1, is fast, intuitive, and automatic, it enables quick reactions based on instinct and experience. The second, System 2, is slower, more analytical, and effortful, used when processing complex or unfamiliar information.
In high-pressure or high-fatigue environments, such as during long shifts or repetitive screening, we tend to rely more heavily on System 1. While this instinctive thinking can be useful, it’s also prone to error. Under strain, we become more vulnerable to confirmation bias (seeing what we expect to see) and satisfaction of search (stopping once one issue has been found, potentially missing others).
But there are moments when intuition proves lifesaving, particularly when it’s questioned rather than blindly followed. In 1986, for instance, an El Al security agent at London’s Heathrow Airport stopped a bomb plot not because of what he saw in an x-ray image, but because something didn’t feel right. Despite the passenger appearing cooperative, the agent’s instinct, reinforced by training and deliberate questioning, led to the discovery of an explosive hidden in the passenger’s bag. It’s a powerful reminder that while our fast-thinking system can signal a potential threat, it’s the slower, more deliberate system that often confirms it.
Interestingly, this blend of rapid decision-making, resilience, and situational control is also found in elite performers, including Olympic medallists. Research shows that top athletes succeed not just through raw ability or training volume, but through a complex interplay of psychological and environmental factors: resilience, motivation, self-regulation, and access to high-quality training environments. These are the very traits security professionals rely on daily, but unlike athletes, they often do so invisibly and without the same systemic support.
Understanding how we think under pressure and designing systems, training, and environments that support both fast and deliberate decision-making is key to improving performance across all areas of security.
Trust, Technology, and the Power of Judgement
As advanced detection tools powered by artificial intelligence (AI) become more prevalent in security operations, the role of the human operator is evolving—not disappearing.
AI systems, such as object recognition algorithms, can flag anomalies with impressive speed and consistency. But in high-stakes environments, they serve best as cognitive partners, not autonomous decision-makers.
The challenge lies in managing this partnership between AI and people. Over-reliance on automation can lead to complacency, while under-trust may cause professionals to disregard valuable insights.
Calibrating that trust requires more than just system accuracy; it demands transparency, training, and the confidence to question both machine output and one’s own instinct. This includes recognizing when a gut feeling warrants a second opinion, a skill seen in practice in the El Al example when a screener, acting on a vague sense that “something wasn’t right,” sought a supervisor’s input. That moment of professional self-doubt, supported by a culture that encouraged escalation, ultimately led to the discovery of a concealed explosive device. In human-machine teams, such calibrated confidence, knowing when to trust, when to question, and when to collaborate, is essential to safe and effective outcomes.
Training for Complexity: Developing People, Not Just Protocols
Far from making screeners obsolete, AI reshapes their contribution. A screener becomes a critical evaluator, interpreting alerts, resolving ambiguity, and applying judgment where machine logic ends. This shift requires a deep investment not only in technical fluency, but in decision-making, attention management, and confidence under uncertainty.
To support this evolution, training must go beyond static instruction. Embedding learning into the operational workflow, through micro-scenarios, real-time feedback, and targeted development, allows screeners to refine skills continuously rather than rely on annual refreshers. When training adapts to the cognitive demands of the job, it becomes a mechanism for professional growth, not just procedural compliance. These efforts aim to produce learning environments that are as dynamic as the threats they prepare screeners to face.
For instance, in a scenario-based training module, a trainee is presented with X-ray images pre-flagged by an AI system trained to detect specific threat categories. One image is marked as containing a possible prohibited item, but the system’s confidence level is low. Instead of accepting or dismissing the prompt at face value, the trainee is guided to slow down, interrogate the image more deliberately, and apply structured visual search strategies. With coaching support, the trainee identifies subtle inconsistencies; for example, an item that mimics a permitted object but has suspicious density patterns.
This kind of targeted training helps new screeners develop not just pattern recognition skills, but also the confidence to question both technology and instinct. Over time, these micro-decisions build the judgement and attention control needed for complex real-world environments where AI support is available, but not infallible. These skills form the foundation of the human–AI teaming that is shaping the future of aviation security.
Security operations increasingly depend on flexible, high-functioning human teams capable of interpreting complex data, managing risk, and acting decisively under pressure. Whether working alongside AI or navigating unexpected scenarios, the screener’s role remains critical. The future of security lies not in replacing human judgement, but in designing systems that amplify it.
Dr. Leanne Simpson is a psychologist with expertise in threat detection, decision-making under pressure, and the psychology of training. With over a decade of experience supporting the defense and security sectors, her work ranges from neuroscience research with the military to the development of evidence-based training for aviation screeners. She currently leads the global training division at Smiths Detection, where she integrates behavioural science with emerging technologies to enhance human performance, psychological wellbeing, and operational resilience at scale.
© Leanne Simpson
