IQSIGHT’s Lewis Stallworth unpacks the nuances of AI-monitored spaces.
If you walk into any modern airport terminal, train station or busy shopping centre today, you only need to look up to spot them.
Sleek, high-definition cameras angled toward the crowd. For the average person moving through these spaces, that sight can trigger an uneasy reaction.
Thanks to decades of Hollywood thrillers and dystopian TV shows, our brains are hardwired to assume the worst.
We look at one of these camera lenses, often enabled with AI capabilities, and imagine a system actively tracking our identities, scanning our facial features or cross-referencing personal details against a database in real-time.
But as someone who spends every day working with this technology, I can tell you that many of these assumptions are inaccurate.
For those of us in the industry, we understand why the public may be sceptical, but the frustration lies in the gap between this perception and actual operational reality.
The most persistent misconception is that every camera is tracking one’s individual movements.
In reality, these environments are less invasive than your everyday afternoon spent browsing the internet on your phone.
To move past the standard privacy debate that happens within our industry, we have to look at what these systems are designed to do – how they analyse patterns, optimise safety and improve the flow of human movement without needing to know your name or face.
What the camera is actually looking at
When an AI-enabled camera observes a transit hub, intersection or retail space, it is not trying to figure out exactly who the individuals are.
It is looking at patterns or metrics that impact the larger, macro-environment of the space.
This includes things like crowd density, dwell times or general behavioural anomalies that can inform public safety and operational decisions on space usage.
To take it a step further, the intelligent sensors are becoming more adept at providing the data security teams need to predict events ahead of time, meaning they can sense and alert potential risks or issues before they actually occur.
To understand how this works in practice, think of the system as an automated note-taker observing the entire scene for patterns.
Instead of capturing a facial biometric signature, the AI-enabled video analytics evaluate size, colour, shape, speed, trajectory and geolocation.
For example, the system could flag slow-moving or stopped vehicles indicating increased congestion, or a crowd formation in a typically free-flowing area.
The data captured can also be analysed to determine where accidents are more likely to happen, based on historical events, to help direct future safety initiatives.
What the public may be surprised to learn is that a massive portion of the public video infrastructure they interact with daily isn’t even recording data.
The majority of infrastructure and transportation cameras are strictly live sensors.
They are deployed to provide real-time incident detection and updates to traffic monitoring centres.
Transit authorities often design these systems as a management and safety tool for traffic management personnel, not to record video.
Why AI is helpful: the operational shift
So, the question becomes, “Why do city leaders need this data if it isn’t for tracking?” The answer lies in shifting the public’s understanding of systems as a tool for enforcement to one that can be leveraged to capture data to inform decisions that help to streamline the flow of people and traffic.
In the public arena, that data limitation directly impacts safety and economic vitality.
Historically, city management and public safety teams deployed assets, such as security staff, based on a map overlay of an area, speculating how crowds would exit a venue or how traffic would back up after an event.
Today, AI-enabled video sensors capture that data for analysis. Those responsible can review data – instead of hours of video recordings – and step into the realm of predictive management.
By recognising patterns over time, authorities can proactively direct traffic, adjust light cycles and deploy safety personnel exactly where the data shows they will best serve the public.
The real-world value of this macro-data is monumental. Take a case study we conducted with one of our city partners.
By using non-invasive AI-enabled data capture to analyse traffic and crowd flow, the city achieved a 4% increase in throughput.
While 4% may sound modest, that increase in logistics and movement translated into an economic revenue increase of nearly a billion dollars for the city.
In retail, operations managers for brick and mortar stores have essentially been running blind compared to their digital counterparts.
I remember having a conversation with an executive who transitioned from an e-commerce giant to a traditional big-box physical store.
Their immediate feedback to us was, “I feel completely blind now.”
Online, they could see where customers lingered, what paths they took and what caused friction.
In the physical world, without AI sensors, mapping those habits and patterns felt like guesswork.
Building the privacy guardrails
To confidently embrace these benefits, leaders must ensure guardrails are established within both the technology and operator process.
At IQSIGHT, we approach this differently by delivering tech that describes an overall event rather than identifying the specific people involved.
To illustrate this, if an anomaly occurs, the system flags data points such as what the person was wearing, what direction they were traveling and the time of the incident.
The system alerts security to the threat or operational hurdle without digging into that individual’s identity or background.
Operational guardrails, which have become a standard practice within the industry, protect data through security measures including encrypted, watermarked video streams and restrictive privacy controls accessible only by administrators.
Additionally, operator activity is fully auditable through the VMS, tracking exactly who looked at what, where and when – ensuring accountability and maintaining public trust.
Closing the perception gap: accountability and open standards
As an industry, opening up the conversation surrounding the negative perception of video surveillance cameras requires a commitment to accountability and open collaboration.
Stakeholders, public space authorities and security directors cannot hide behind good intentions. Instead, they must actively take a role in educating the public on what their infrastructure is really doing.
Accuracy and transparency matter. AI-enabled features are a mechanism to understand, protect and optimise our shared spaces, not spy on individuals.
By keeping our systems open, our data auditable and our focus squarely on macro metrics, we can build an environment that is remarkably efficient and safe – all while keeping personal privacy intact.
