We close out the year with 2025 security industry reflections and 2026 predictions from Sal Mani, co-founder and head of safety and security at VOLT AI. He looks back at the year that was in the physical security space and ahead to what changes could be in store for the next 12 months and beyond.
Security Sales & Integration: What has been the most surprising development in the security industry this year?
Sal Mani: The most surprising development in the security industry this year is that artificial intelligence quietly stopped being “just analytics” and became the operating system of physical security. For decades, we argued over camera brands, badge types and VMS features, and now the real power has shifted to the software brain sitting on top of it all.
In 2025, AI is not just drawing boxes on video; it is triaging alerts, drafting incident summaries, guiding operators through playbooks, coordinating robots and drones and squeezing new value out of old cameras that should have been ripped out years ago.
Security teams that once treated AI as a lab project or a bolt on module are now budgeting for it as core infrastructure, right alongside access control and networks. That pivot from hardware-obsessed to AI-driven, from “extra license” to mission-critical, happened faster and more deeply this year than most of us expected and it is reshaping what it even means to run a modern security program.
SSI: What technology category or solution area do you see as 2026’s ripest, most profitable growth opportunity for security dealers, installers and integrators? Explain your reasoning.
Mani: The strongest growth opportunity for 2026, in my view, is cloud-based managed security services: delivering video, access control, system health and analytics as an ongoing service rather than just installing hardware and walking away.
Hardware margins keep getting thinner, ownership groups are pushing for recurring revenue instead of one time projects, and customers are increasingly buying outcomes instead of devices. They want uptime guarantees, usable alerts instead of noise, help with compliance and reporting and a partner who will actively manage and tune their systems over time.
The integrators who design and deploy the stack, then stay in the loop to monitor, optimize and report on performance month after month, will capture far more long-term value than those who simply compete on camera price. This shift from project work to service relationships, wrapped around AI and cloud, is where the real profit and stickiness will be.
SSI: Which emerging security technologies do you think are overplayed? Which ones do you think will truly transform the practice of security integration in the coming years?
Mani: Some of the most overplayed technologies right now are standalone facial recognition, point solution AI widgets and gadget-heavy IoT that lacks a real platform. Facial recognition gets a ton of attention but, in most real environments, the privacy, bias, legal and public perception issues make it a political project more than a practical one.
Single-purpose AI boxes for one niche use case, like a slip-and-fall detector bolted onto the side, create extra user interfaces and support headaches without integrating cleanly into the customer’s incident management protocols and tools. Random smart devices that do not speak open standards or push usable data into a central system become technical debt the integrator has to babysit, not a scalable business.
The technologies that will truly transform security integration are open, API-driven platforms with AI built in, AI-assisted incident management and automation, robotics and drones tied directly into command workflows, digital twins and 3D mapping and identity centric physical access.
Open cloud or hybrid platforms with native AI detection and triage turn integrators into designers of data flows, automations, and custom apps, which is much higher margin than simply installing boxes. AI-assisted incident tools will sit across video, access and sensors to triage events and guide operators through playbooks, which opens the door to continuous optimization and service based engagements.
Robotics and drones become powerful when they are treated as first response sensors inside the normal VMS and dispatch process, not toys. Digital twins let you design, simulate and operate sites from a single 3D model that carries through from pre-sales to the SOC.
And tying physical access into IT identity and zero trust means badges, HR systems, and directories all work as one fabric, opening the door to ongoing governance and lifecycle management as a service. That combination is what will change how integrators design, bill and stay indispensable to their customers.
SSI: What’s getting better about the security industry these days? What seems to be getting worse?
Mani: What is finally moving in the right direction is how we think about privacy and how AI supports people on the front lines. Privacy is no longer a box to check, it is becoming the baseline test for whether an AI security system even belongs in the conversation.
Doing this right means focusing on behavior, not appearance, designing systems that do not need to store personally identifiable information to be effective and keeping a human in the loop when decisions affect people’s lives, jobs or reputations.
When you build AI this way, it does not replace your security staff; it amplifies them. It gives them better visibility, cleaner signals and faster context so they can do the job they were hired to do with more confidence and less guesswork.
At the same time, AI-driven hoaxes like swatting are getting more sophisticated and more frequent and they will follow us into 2026. Every school, campus and organization has to accept a hard truth: a false alarm can be just as disruptive and traumatic as a real event. This is exactly where AI should act as a force multiplier, not an autopilot.
Human-validated AI can rapidly cross check video, audio and access data, flag inconsistencies and help security teams rule out hoaxes before panic spreads. The human in the loop becomes the filter that reduces noise, stops alert fatigue and makes sure that when an alert goes out people can trust it is grounded in real evidence, not manufactured drama.
SSI: What’s the single most pressing challenge that professionals in the security industry must tackle right now? And how would you suggest tackling it?
The most urgent challenge in front of us is not adding more cameras or more alerts; it is separating real danger from increasingly polished hoaxes before they spin a community into chaos. Swatting is the clearest warning shot. AI is about to pour fuel on it.
Bad actors can already create convincing gunshot audio, clone voices for emergency calls and script armies of bots to flood 9-1-1. The cost is low, the tools are easy to get and the impact is a school, a campus or an entire city yanked into crisis mode on command. That is the reality security leaders and first responders are walking into in 2026.
The only credible way to meet that threat is to use AI on the defensive side while insisting that human judgment stays in control. AI should be the engine that fuses video, audio, access control and historical patterns into fast, usable insight so humans can make better calls with more context and less guesswork.
It should help rank credibility, cross-check signals and compress time, not replace the person who decides what is real and what is not. Keeping a human in the loop is not a nice to have, it is the safeguard against bias, false alarms, overreaction, and ultimately the erosion of trust in the systems we are asking people to rely on.
The way to tackle this is pairing AI speed with human judgment. AI should give security teams faster insights and stronger context so humans can make smarter calls. But it should never operate without human validation — that’s the safeguard against bias, false alarms, and misuse.
SSI: Finish this sentence: 2026 will be remembered as the year that the security industry…
Mani: 2026 will be remembered as the year the security industry flipped the script, breaking free from box selling and becoming a force of intelligent, easy to use, always on protection powered by AI, cloud, and deeply integrated, outcome driven services.
The post Sal Mani, VOLT AI: 2025 Security Reflections and 2026 Predictions appeared first on Security Sales & Integration.
