The AI Transformation in Security Integration

Like many tech-driven industries, the security industry is undergoing a massive transition – most of which is powered by Artificial Intelligence. This is not breaking news; after all, in 2024, the Security Industry Association devoted the first four of its 10 annual “megatrends” to AI.

But while it isn’t breaking news that the industry is adapting to the use of AI in technology, the executives running security integration and consulting firms have probably felt like a leaf in a windstorm, being tossed about by what seems like daily technology innovations that greatly impact both their customers and their own businesses.  

It isn’t enough to know how AI will work for video analytics; now they need to know how it works across the gamut of all security and even non-security technologies. Even more daunting, these seasoned executives are now being told that failure to implement AI from an internal operations basis means they are behind the curve – a sure path to eventual perish.

The efficiencies created by AI makes things faster, but more importantly, it is giving human talent the bandwidth to focus on higher-value work.

“The routine, low-knowledge, low-skill tasks that we just have to do on a day-to-day basis are the greatest opportunity [for AI],” Castillo says. “That gives us back time so that our folks are not using their powerful minds on routine things that require little skill.”

Adds Nemerofsky: “Now our coordinators can be more customer-focused and more employee-focused, doing the A-type work we wanted them to do.”

From Service Provider to Strategic Partner

Beyond the back office, as customer interest in AI functionality piques, it has moved from a pie-in-the-sky idea to meaningful purchase orders. This means integrators must fundamentally change how they engage with their clients to create deeper, more consultative relationships.

Castillo observes a clear market shift: “What was just interest a couple of years ago has now turned into purchase orders,” he says. “Most of our customers – primarily in the enterprise space – now understand the force multiplier that AI can be, and they want to leverage it.”

Leveraging it for customers means understanding the greater impact of the technology; in fact, the AI technology’s significance to the security department may actually become secondary to the impact it has on an overall enterprise operation. This means Castillo must “get into a deeper relationship with customers to understand their business more holistically than we ever have – because it is not just a security system, it is a business system that customers can leverage to make money, save money, and reduce risk.”

In turn, this deeper engagement extends beyond traditional security stakeholders. “I love to at least engage or invite to the table folks from sales, marketing, and operations and explain how they can leverage the investment in the security system,” Castillo adds. “It ends up being a win-win.”

For integration company staff experts, the consultative approach is paramount for AI evaluation and deployment. “The average clients who don’t have a technology lead are overwhelmed by options,” Nemerofsky says. “Our role is to be a guide or an architect on how this is going to look, how it aligns with their tech stack, and dealing with the IT environment and with business objectives. It’s not just something new and shiny.”

SAGE and Nemerofsky champion the embedded model for integration for this very reason. “Having embedded people is a huge advantage, because you understand your client’s tech stack better than they understand it, or at least as well as they understand it,” Nemerofsky says. “You are better able to be a resource and a consultant to a client when you are embedded.”

Technology Vetting and Quality Assurance

With numerous AI solutions entering the market coming from all means of manufacturers – from the start-ups to the well-established – integrators have been forced to develop sophisticated evaluation processes that extend to all new technologies. “Our review team evaluates three technologies a week during a typical month,” Nemerofsky says.

The vetting process must be comprehensive. “There’s an initial evaluation, where a vendor shows us what it does in their environment,” Nemerofsky explains. “Next, we will put it in our dev environment and test it on, say, 15,000 cameras. If it works in our dev environment for 90 days, then we will bring it to test for the customer. Rarely do we go from dev to deployment for an entire enterprise at one time.”

Castillo takes a similarly thorough approach at PrefTech: “We have an innovation team that does due diligence,” he explains. “They analyze the company itself, then they look at the technology and test it. We look at everything from support after the sale to training tools and classes they have.”

The complexity of AI solutions has made proof of concepts standard practice for enterprise deployments, and forward-thinking integrators are helping larger customers establish their own testing capabilities. “We recommend our enterprise customers create a mechanism to test,” Castillo says. “Building out these pilot facilities and test labs may cost $50,000 or $100,000, but it could save millions in bad decisions.”

The Vendor Ecosystem Transformation

AI is reshaping the traditional vendor-integrator-customer relationship chain. “We’ve moved beyond transactional vendor relationships,” Nemerofsky says. “Instead of just buying boxes, we expect our technology partners to deliver outcomes for our clients – and that means transparency and giving us a view into the product R&D roadmap.”

Meanwhile, many of the newer AI-focused market entrants in the vendor space are bypassing traditional sales channels, Nemerofsky says. “They aren’t coming in and selling through distributors like PSA or Wesco or through integrators – they are going right to the end-user. Some of these new AI entrants have been clever enough to create rebate programs for the integrator; others are just selling it directly [integrators] are cut out.”

This creates unnecessary tension between the integrator/consultant and the customer when the integrator gets a sudden phone call from a customer about a cool technology they were introduced to. “One of two things will happen,” Nemerofsky says. “Because they can buy it directly, they want to know how we are going to integrate it into their tech stack; or, it is a big Fortune 500 company that takes so long to onboard a new technology that they push it back through their integration channel, to us.”

Despite the channel disruption, AI is fostering deeper relationships between integrators, customers, and vendors. “There’s more collaboration between the manufacturers and security integrators, and the end-users,” Castillo confirms. “We are solving problems together, because hopefully the manufacturer knows the product better than we do.”

Much like the business back office changes, the business relationship with vendors has adjusted with the introduction of AI tools. “[Vendors are creating ways] to leverage AI tools to make buying easier, tracking easier, updates easier, and they are still finding new ways to improve that.”

Innovation vs. Stability

Not all markets are ready to embrace cutting-edge AI – particularly government and regulated sectors, which make up the lion’s share of Lanning’s customers. This makes it much more difficult to be on the front lines, testing and vetting new technologies like Castillo and Nemerofsky do.

“When you deploy a software program on a government network, you need what’s called an ‘authority to operate,’ and it can take literally up to 24 months to get one,” Lanning says. “You can imagine if I present the new whiz-bang version of Lenel software [to a government client], it’s going to take 18-24 months to get that acceptance, and by that time, Lenel has already released two newer versions.”

This has forced a conservative technology approach upon Lanning; in fact, while many of us are marveling at the cool innovations presented in tiny 10x10 booths in the back corner at ISC West, she says she refuses to even check them out.

“We tend to stick with trusted products that have been around a while,” she admits. “I don’t even go in that whole section [at ISC West]. Because of our government requirements, we simply can’t look at bleeding-edge technologies. I would just say that our customers are willing to sacrifice innovation for stability when it comes to security systems.”

Perhaps that’s why Lanning sees limited practical AI implementations in spite of all the hype. “A lot of the AI functionality, in my opinion, is just not there yet,” she says. “I see people experimenting with some stuff, but again, most of our customers are super vanilla.”

Workforce Evolution in an AI World

Like it or not, AI is driving fundamental changes in workforce and technician requirements, and end-user customer expectations are driving them.

Far beyond the traditional security disciplines, Castillo emphasizes that this new environment requires comprehensive technical capabilities. “We believe it’s an imperative that modern security integrators are competent with the entire tech stack – from workstations to server storage, networking, operating systems, cybersecurity tools, software, and obviously applications,” he says.

“Training has become a huge investment for the integrator,” Nemerofsky says. “You can’t just train them on the product that you’re deploying; they’ve got to be cyber certified, and they’ve got to have network and cloud platform training, along with an understanding of traditional cabling. 85% of our people are network satisfied and technology satisfied, or certified. You’ve got to build them along that career path for a long time. That’s why you hear me being such a proponent for FAST, the Foundation for Advancing Security Talent.”

The AI Vision Moving Forward

Despite current limitations, industry leaders – from integrator to consultant to end-user – have clear expectations for valuable AI. “We need real case studies where it solved a problem, because AI for AI’s sake is useless,” Lanning says.

Castillo sees particular promise in response systems: “We’re very good at detection in our industry, but automating responses – especially in emotional situations and helping operators get through workflows when you’re not thinking this clearly as they should – would be a great use for AI.”

In the end, whether it is for a customer’s practical application or for improved internal business practices, today’s (and tomorrow’s) successful integrators will embrace AI as a tool for operational efficiency while maintaining focus on customer outcomes and system reliability.

“The practitioners have become very smart,” Nemerofsky says. “They are smart evaluators, smart buyers, and they know the right questions to ask and how to get to the end-result they want. So we need to be just as smart.”