The United States operates roughly 440 commercial airports, and nearly every square foot of them sits under a security camera. Yet on January 6, 2017, a gunman retrieved a pistol from his checked bag at Fort Lauderdale-Hollywood International Airport and killed five people in the baggage claim area, a zone fully covered by airport security cameras that recorded everything and detected nothing. The attack happened on the landside, the public half of the airport where federal screening does not apply, and it remains the clearest illustration of a gap most airport security plans still have not closed.
Airport security cameras are everywhere, but at most facilities they remain passive recorders. This article looks at where the landside gap comes from, what the federal data says about it, and how AI video analytics turns the cameras airports already own into a real-time detection layer, with no facial recognition and no camera replacement.
One Airport, Two Security Realities
Every commercial airport is split into two security environments: the airside, where TSA screens every passenger and bag, and the landside, where anyone can walk in carrying anything. Past the checkpoint, aviation security is among the most rigorous in the world. TSA screened more than 900 million passengers in 2024 and stopped 6,678 firearms at checkpoints that year, 94 percent of them loaded, according to TSA's January 2025 reporting.
Before the checkpoint, none of that protection exists. The departures curb, the parking garage, the ticketing lobby, and baggage claim are open to the public around the clock. They are the responsibility of the airport operator and local law enforcement, not TSA. The Fort Lauderdale shooting happened there. So did the 2013 LAX shooting, in which a gunman walked through a public terminal entrance and killed a TSA officer before ever reaching a screening lane. The landside is where crowds are densest, screening is absent, and cameras are most likely to be unmonitored.
The core problem: the most heavily trafficked zones of an airport sit outside the federal screening perimeter, protected mainly by cameras that record incidents instead of detecting them.
What the Federal Data Says About the Gap
U.S. airports averaged roughly 2,500 perimeter and access-control security incidents per year between fiscal years 2009 and 2015, according to TSA data analyzed by the Government Accountability Office. The GAO's 2016 audit of airport perimeter and access control security counted between 2,200 and 2,800 incidents annually, found the trend moving upward, and noted that TSA had analyzed breach data at only 19 percent of the nation's commercial airports. More than 1,670 breaches at small and midsize airports were never analyzed at all.
Those figures are fence jumpers, vehicle gate runners, and unauthorized access through employee doors, the exact events perimeter cameras exist to catch. As of the GAO's review, the cameras caught them on video. They just did not catch them in time. The same pattern shows up across critical infrastructure, which is why AI perimeter security has become the standard answer to fences that record breaches instead of stopping them.
Incidents keep making the point. In January 2025, a driver at Appleton International Airport in Wisconsin ignored multiple warnings, drove through an open construction gate, and nearly reached an active runway before being stopped. The vehicle crossed camera views the entire way.
Why Passive Airport Security Cameras Fail
Airport security cameras fail for the same reason cameras fail everywhere: humans cannot monitor hundreds of feeds, and unmonitored video only matters after the incident is over. A large hub airport can run several thousand cameras across terminals, garages, curbs, and fence lines. No operations center staffs enough people to watch them, and research on vigilance consistently shows that human attention to monitor walls degrades sharply within about 20 minutes.
The result is a security posture that is reactive by design. Cameras provide evidence for the investigation, not warning for the response. Airport police are dispatched when someone calls 911, which means the response clock starts minutes after the threat became visible on camera. Public transit agencies face the identical problem, and the identical math: every major U.S. transit system has cameras, and almost none of them can detect a weapon.
What AI Video Analytics Changes, Zone by Zone
AI video analytics converts existing airport cameras from passive recorders into active sensors that flag threats within seconds and route alerts to the people who can respond. Instead of waiting for a human to notice, the AI watches every feed at once, around the clock, and escalates only what matters. Here is what that looks like across the landside zones the checkpoint never touches.
| Zone | Typical risk | AI detection layer |
|---|---|---|
| Departures and arrivals curb | Brandished weapons, vehicle-as-weapon approaches, unattended crowding | Real-time weapon detection, vehicle detection in pedestrian zones, crowd formation alerts |
| Parking garages and lots | Assaults, theft, loitering, medical emergencies out of sight | Loitering detection, fall detection, weapon detection on existing garage cameras |
| Ticketing lobby and baggage claim | The Fort Lauderdale scenario: an armed attacker in an unscreened crowd | Weapon detection that alerts security within seconds, before the first 911 call |
| Perimeter and fence line | The roughly 2,500 annual breaches GAO documented | Perimeter and rooftop intrusion detection with immediate alerts, not morning-after video review |
| Employee access doors | Tailgating and unauthorized access to secure areas | Unauthorized access detection layered on the cameras already covering each door |
| Terminal interiors | Slip and fall injuries, smoke, unfolding disturbances | Fall detection, slip-and-fall hazard identification, visual smoke and fire detection |
The vehicle threat deserves emphasis. Curbside zones concentrate hundreds of pedestrians a few feet from open traffic lanes, and vehicle ramming attacks are surging while passive cameras can do nothing about them. AI that recognizes a vehicle entering a pedestrian zone buys the seconds that bollards and barriers alone cannot.
No Facial Recognition Required
Effective airport threat detection does not require identifying a single traveler. That distinction matters more at airports than almost anywhere else, because biometric screening is already one of aviation's most contested privacy debates. Several legacy vendors lead their airport offerings with facial recognition and watch lists, which drags airport operators into surveillance policy fights, biometric consent requirements, and state privacy statutes.
Threat-based AI sidesteps all of it. The system detects objects and events: a drawn weapon, a person on the fence line, a vehicle where vehicles do not belong, a traveler collapsed in a garage stairwell. It does not identify people, store biometric profiles, or match faces against databases. IntelliSee's guide to AI gun detection covers in depth how visual weapon detection works without identifying anyone, and why that architecture clears privacy review faster, a meaningful advantage for public-sector airport authorities answerable to elected boards.
Key takeaways for airport security directors
The landside is the soft target: federal screening never touches the curb, garage, lobby, or baggage claim. GAO documented roughly 2,500 perimeter and access incidents per year as of its 2016 audit, and passive cameras only record them. AI video analytics layers real-time weapon, perimeter, vehicle, crowd, loitering, and fall detection onto the cameras an airport already owns, with no facial recognition and no rip-and-replace.
Deploying on the Cameras You Already Own
The fastest path to landside detection is software, not construction. Airports have spent decades building out camera coverage; the infrastructure investment is already made. An AI risk mitigation platform like IntelliSee layers onto that existing camera network, analyzes the live feeds in real time, and pushes alerts to security operations, mobile devices, and integrated systems within seconds of a detection. There is no hardware replacement, no proprietary camera requirement, and no need to rebuild the operations center.
That deployment model also changes the budget conversation. Capital-intensive proposals, new camera networks, additional staffed posts, or screening infrastructure on the landside, compete for funding measured in years. A software layer on existing infrastructure is an operating decision, and for public airport authorities, eligible projects can often tap security-focused grant streams. IntelliSee's grant funding hub tracks the federal and state programs that help public agencies fund AI video analytics.
IntelliSee's platform carries full DHS SAFETY Act QATT designation, the same federal designation tier held by the largest names in weapons detection, and is deployed today across K-12 districts, hospitals, corporate campuses, and public venues facing the same problem airports face: too many cameras, not enough eyes.
Frequently Asked Questions
Do airport security cameras detect weapons?
Standard airport security cameras do not detect weapons; they record video for later review. Weapon detection requires either screening equipment at checkpoints or AI video analytics software that analyzes camera feeds in real time and alerts security within seconds of a visible weapon.
Who is responsible for landside airport security?
Landside areas such as the curb, parking garages, ticketing lobbies, and baggage claim are secured by the airport operator and local or airport law enforcement, not TSA. Federal screening responsibility begins at the security checkpoint.
Can AI video analytics run on existing airport cameras?
Yes. Modern AI video analytics platforms layer onto existing IP camera infrastructure and analyze live feeds in real time, which means airports can add detection capability without replacing cameras or rebuilding their video management systems.
Does AI threat detection at airports use facial recognition?
It does not have to. Threat-based platforms like IntelliSee detect objects and events, such as weapons, perimeter intrusions, vehicles in pedestrian zones, and falls, without identifying individuals, storing biometric data, or matching faces against watch lists.
The Cameras Are Already There. Give Them a Job.
Airports are among the most camera-dense environments in the country, and among the most studied soft targets. The federal data has been public since 2016. The incidents keep arriving on schedule. What has changed is that the passive camera network covering every curb, garage, and baggage carousel can now become a proactive detection layer in a software deployment, not a construction project.
IntelliSee's mission is turning passive cameras into proactive protectors. If your airport, authority, or port district is evaluating AI video analytics for landside security, talk to our team about what your existing cameras can do, and visit the grant funding hub to see which programs can help pay for it.