Editor’s note: FM Perspectives are industry op-eds. The views expressed are the authors’ and do not necessarily reflect those of Facilities Management Advisor.
Twenty-five years ago, you could walk your daughter to her airport gate when she left for college. Families waited at the jet bridge to say goodbye. In office towers, anyone could walk onto almost any floor. A school’s front entrance was often just an unlocked door. Few Americans gave the security of these places a second thought. September 11, 2001, changed that.
The Security Gap We Accepted
In the years that followed, the U.S. made an enormous investment in physical security, particularly at airports and other high-profile targets. Aviation security spending alone exceeded $5.8 billion in fiscal year 2002, compared with roughly $1 billion a year before the attacks. By 2005, the federal government’s homeland security budget had more than doubled to $54.4 billion, with an estimated $35 billion of that increase attributable directly to 9/11.
Airports could justify sophisticated screening, surveillance, and monitoring because the stakes were extraordinarily high. The same was true for government buildings, critical infrastructure, and some of the world’s largest commercial properties. Everywhere else, the economics didn’t work.
Most schools, stores, factories, healthcare facilities, warehouses, apartment buildings, and businesses couldn’t afford teams of people watching cameras around the clock, systems analyzing every access event, or the infrastructure required to run it all.
So, they settled for a different model: a camera by the door, a badge reader, someone to call when an alarm went off. It wasn’t that these places didn’t want airport-level security. They couldn’t justify the cost.
That is the gap artificial intelligence (AI) closes. AI does more than make security smarter. It makes a level of security once reserved for airports economically viable across millions of other places.
From Recording What Happened to Understanding What Is Happening
The first transformation in physical security was the camera. Analog systems gave way to IP cameras, which gave way to cloud-connected ones that were easier to deploy, update, and manage. For most of the past two decades, cameras remained primarily forensic tools: Something happened, and later, someone reviewed the footage to figure out what.
AI-enabled cameras flip that. An intelligent camera can recognize events and patterns as they happen, giving people a chance to respond before a situation escalates. It can flag someone entering a building at an unusual hour, an unauthorized person moving into a restricted area, a worker not wearing required safety equipment, or when a person is brandishing a gun, and get that information to the right person fast enough to matter.
The difference goes beyond detection. It’s scale, and that’s what changes the economics. A person can watch a handful of screens. AI can watch thousands of cameras continuously and surface only what’s worth a person’s attention, at a fraction of the cost of security guards watching a wall of monitors around the clock.
Access Control: Asking Better Questions
Access control is going through a similar shift. For decades, a badge or key fob answered one question: Does this credential open this door?
An intelligent system can ask much more useful questions: Does it make sense for this person to enter this location right now? Is the badge being used at an unusual hour, accessing a location the person has never entered before? Does it fit that person’s typical pattern?
An unusual event doesn’t necessarily mean something is wrong, but it can be a reason to take a second look. That kind of review used to require a security team combing through access records by hand, a cost most buildings could never justify. AI makes the same level of oversight available at a fraction of the price.
One System, Unified
We’re now in the next phase: unifying what used to be separate systems. We can bring video surveillance, access control, visitor management, and intrusion detection together in one place, and connect them to a wide range of third-party technologies and data sources instead of operating as isolated boxes. A camera shouldn’t have to work in isolation from a badge reader; a badge reader shouldn’t be isolated from the door it controls, or from the identity of the person using it.
When video, access, sensors, and identity converge around a shared context layer, the system starts to understand when a person, a credential, a door, a camera, a vehicle, and an incident are connected. A notification that says “someone is in the parking lot” becomes an alert that identifies an unauthorized person entering a restricted area, ties that event to the access control record, and automatically pulls up the relevant video for a security operator.
That’s also why the competitive ground in this industry is shifting: It used to be about hardware, better cameras, better readers. Increasingly, it’s about who can combine the most useful data, integrations, and workflows into a system that operators trust.
That is where AI becomes more than another feature on a camera. It becomes the intelligence layer across the security environment.
Cloud Makes the Economics Work
None of this happens at the scale that matters without the cloud. Most companies moved their core business systems—email, CRM, accounting—to the cloud years ago because it’s easier to manage, update, and secure. Physical security followed later, with adoption accelerating in earnest between roughly 2017 and 2020.
Novaira Insights, which tracks the video surveillance market, has watched that shift take hold: It forecasts the number of cloud-connected cameras in the U.S. to keep growing at roughly 80% a year and reported more than 2 million professional-grade cloud-connected cameras already in place in the U.S. by the end of 2022. Most organizations upgrading their security systems today choose cloud-based ones by default, the same way they would for any other business system.
The economics show up in real deployments, not just industry-wide forecasts. One national multifamily operator managing more than 10,000 doors across dozens of properties cut its per-building engineering costs from roughly $6,000 a year to about $1,500 after moving from legacy hardware to a cloud-based access platform, largely by eliminating the truck rolls that used to be required every time a lease turned over or a lock needed to be reset. Across its full portfolio, that added up to more than $150,000 in annual savings.
Cloud is what makes AI affordable to run at scale, and AI is much of what makes the cloud connection worth having in the first place. Neither gets you to airport-level security on its own. Together, cloud and AI make it affordable.
A Smarter System Still Needs People
Responsible deployment is critical. Video data is sensitive, and organizations have an obligation to protect it and use it deliberately. AI shouldn’t be the one making the judgment call. Its job is to surface what a camera or badge reader captures so a human can decide what happens next, faster.
That’s really the point of the economics, too. AI doesn’t replace the person making the decision. It makes it possible for that person to watch over far more ground than any budget could previously cover.
Airport-Level Security, Without the Airport-Level Budget
After 9/11, the country made a massive investment in securing the places most directly tied to the attacks. That investment changed aviation security and reset the country’s expectations for what physical security should look like. The economics of AI means we can now cost-effectively extend that thinking beyond airports.
AI can make security more intelligent. Cloud infrastructure can make it more practical to deploy and maintain. Together, they can make a higher level of protection economically feasible in the millions of places where airport-style security was never realistic.
For 25 years, we’ve accepted that sophisticated physical security was something only certain places could afford. AI challenges that assumption.
Dean Drako is chairman and CEO of Brivo, a provider cloud-native AI-driven physical security solutions. In 2012, Drako founded and served as CEO of Eagle Eye Networks, which he merged with Brivo in 2025. He is also chairman of enterprise security automation company Cobalt AI, as well as founder and CEO of electric vehicle company Drako Motors. Previously, Drako was founder, president, and CEO of Barracuda Networks. He received a B.S. in electrical engineering from the University of Michigan and an M.S. in electrical engineering from UC Berkeley.