Going The Distance with Agentic AI on the Edge

Agentic AI enables cameras to independently analyze threats, make decisions, and trigger responses locally.

• By Monique Merhige
• Sep 17, 2026

In today's world, Artificial Intelligence is accelerating like a race car on an open road, leaving many companies in the dust, struggling to keep up with nonstop technological advances.

Organizations are feeling increased pressure to integrate AI into their existing security platforms and are turning to their trusted security partners for guidance. Unfortunately, the industry is working hard to catch up and help safeguard those investments.

Over the years, security has seen its fair share of transformations – from analog to IP video, from on-prem to cloud, and now from simple chatbots and visual LLMs that identify objects to Agentic AI that can reason and think. Unlike traditional AI, Agentic AI can observe an environment, make decisions, take actions, learn from outcomes, coordinate with other agents, and operate continuously with minimal human input. Understanding its role and benefits should be a top priority for every security professional.

The Rise of the AI Revolution

ChatGPT kicked off the AI revolution in 2022. By 2024, the momentum accelerated when the year became defined by prompting and chatting. In 2025, it evolved in two significant ways. First came the rise of Model Context Protocol (MCP), which is the ability to plug tools into AI and put it to work, whether generating presentations or pulling real-time third-party data.

Second came Retrieval Augmented Generation (RAG), which allows AI models to answer questions based on your organization's own documents rather than general training data. Imagine asking an AI about your company's HR policies and receiving answers drawn directly from your employee handbook. There are endless ways to leverage AI, but understanding the best applications for your specific security and operational needs is what separates leaders from laggards.

Expert Insights

An Interview with David Antar and Gus Spathis of ViewScan
To gain deeper insight into computer vision and Agentic AI, it helps to hear from true industry experts. David Antar, president of ViewScan, is a serial entrepreneur who was the founder of IPVideo Corp, co-invented the Halo IoT Smart Sensor, and founded A+ Technology & Security, which was acquired by Convergint.

He has been inducted into the Long Island Technology Hall of Fame and has received numerous honors, including the Visionary Award from the Long Island Software and Technology Network (LISTnet) and Industry Partner of the Year at NYIT's Ingenium Awards. He is deeply involved with local universities and follows his passion for making the world a smarter, safer and healthier place.

Gus Spathis, strategic partner at ViewScan, is an experienced engineer and internet pioneer with more than 20 years in technology. He has created award-winning solutions across multiple industries and has been involved with machine learning and AI innovation since the 1990s, consistently pushing the boundaries of what intelligent systems can achieve.

The two partnered in early 2025 after serving together on the boards of Hofstra University and LISTnet. "Long Island has become a mini-Silicon Valley," Antar said, "with prestigious universities like Hofstra, SUNY Farmingdale, NYIT, and Stony Brook driving collaboration among technology-minded people. We are excited to help organizations around the world give life to their existing security investments and bring intelligence to the edge."

The Partnership Pushing AI Forward

Can you explain your roles and mission at ViewScan?
Antar. “ViewScan delivers a modern physical security platform by unifying video surveillance, access, and sensor technologies with the real-time intelligence of mission-oriented remote security operators, primarily disabled American Veterans. With prior military training, they understand what a mission entails.

“Our software ecosystem is built on three core components: an AI Engine, a series of advanced unified models delivering multiple analytics on any video stream without specialized hardware; an Event Manager that continuously analyzes video and sensor signals and generates real-time alerts; and front-end software that creates a true "single pane of glass" for situational awareness — usable by our operators, the end user, or a hybrid of both. Gus ensures every component works seamlessly together, applying best practices to leverage the right technology in the right way. Our goal is simple: our technology should save you money, increase operational efficiency, and create a safer environment. If we do not accomplish that, we have not met our mission.”

Spathis. “Just like a conductor guiding an orchestra, my role is to help ViewScan harness the latest AI technologies to build the most forward-thinking and sustainable security platforms in the industry. In an industry where today's innovation can quickly become yesterday's news, we stay relentlessly focused on bringing the most advanced, practical AI capabilities to our clients.

“What we are doing on the edge is different. We are applying the latest breakthroughs in computer vision, making our own fundamental new advancements, and packaging it all to operate at the edge. This means organizations can unlock more value from the systems they already own, eliminating the need to rip and replace anything, while reducing operational costs and enabling a smarter approach to security.”

Agentic AI in Action

How is Agentic AI Impacting the Security Industry?
Antar. Security cameras are no longer about preventing theft. Agentic AI can help determine buying patterns in retail, monitor occupancy levels in real time, and drive operational efficiency across every vertical market. The key is understanding agentic security camera workflows.

“Each workflow is an AI "agent" attached to a fixed camera. An agent is simply software running a repeating loop: it watches a video feed, decides whether something matters, acts if necessary, and remembers what happened so it gets smarter over time. A security guard does not need to stare at a monitor for it to work. The same four-step pattern drives every use case with only the trigger and the response changing.

Spathis. This is the year of the agent, where agentic workflows are transforming how the security industry operates. Best practice calls for Humans-in-the-Loop (HITL), both in real time during incidents and offline, to review image data used to train the next generation of models. HITL is the method where humans provide feedback, correct errors, or make final decisions to enhance accuracy and reliability. Together, human oversight and agentic intelligence create a security ecosystem that is smarter, faster, and more adaptive than anything we have seen before.

Here is a real-world example: Perimeter Watch. A camera checks a fenced electrical yard at night. It watches for anyone lingering near the razor-wire fence for more than two minutes. It thinks — considering whether the person is just passing through, carrying tools, or whether any overnight maintenance has been scheduled. If no work is scheduled and the person approaches the gate, it plays a recorded "restricted area" audio warning. If they begin climbing the fence, it dispatches police immediately.

Where Does Agentic AI Stand Today, in 2027 and Beyond?

Antar. There's AI — and then there's Agentic AI at the edge, the wave of the future. Much like Tesla transformed real-time decision-making into vehicles, we are enabling organizations to rely on live, local data rather than the internet to drive intelligent business decisions. We are thrilled to be at the forefront of this shift.

Spathis. Agentic workflows will continue gaining traction into 2027 and beyond as organizations push intelligence closer to the edge. We will see a strong migration away from cloud-dependent solutions turning back towards on-prem devices infused with intelligence. Managers will also be able to leverage their internal teams far more effectively as Agentic AI takes on routine tasks, streamlines workflows and frees up resources to focus on higher-value work.

Agentic AI: The Engine Behind Future Transformation

AI has taken the security industry on a remarkable ride over the past few years. Like any meaningful journey, it comes with twists, bumps and sharp turns, but those challenges are exactly what shapes the path forward.

As the landscape matures, the road ahead will become smoother with the help of Agentic AI and edge-driven technology. By giving organizations, a clearer path towards stronger security, higher efficiency, and real-time intelligence, we are fueling their internal engine by empowering AI to think, act, and optimize alongside human teams, and not replace them.

This article originally appeared in the September/October 2026 issue of Security Today.