AI: Filtering through the noise


sh Risby MSyI , Chair of the Security Institute AI & New Emerging Security Technologies Special Interest Group, has developed the FILTER approach to help you decide whether AI is right for your organisation and shows some simple steps to get started.


A


Like many of the technologies we’re seeing enter the security and defence markets, the focus of my article is ‘dual use’ in nature. That is, using AI to filter through noise, but also understanding the noise associated with AI adoption itself.


You’d have to have been living under a rock not to appreciate the immense opportunities AI offers humanity – from identifying a suspect on a watchlist within a crowd to notifying your smart watch of that gone-off milk in your fridge, the spectrum of use cases is broad. From the ground-breaking and novel to the trivial, AI is proliferating in daily life whether we like it or not.


None more so than in security where AI is now the go-to buzzword for any sales executive trying to flog the latest and greatest security innovation. But are we adopting AI for AI’s sake – and do we even know what we’re signing up to? Everyone's transmitting on the AI frequency right now. The harder question is whether you're actually receiving anything useful?


Used correctly, AI can revolutionise how we approach all aspects of security, providing meaningful operational efficiencies, resource utilisation and improved client outcomes. But how do we deliver this practically – where do we even start?


To help with this, I’ve developed a FILTER to help you decide whether AI is right for your organisation and I show some simple steps to get started:


F – Find the Gap


Before you look at a single supplier website or brochure, be honest with yourself about where your security operations actually have gaps. What's slow, inconsistent,


resource-intensive, or error-prone? Then ask the harder question: does fixing that genuinely require AI, or would a straightforward change to your current processes or procedures deliver the same result?


It sounds obvious, but it's a step that gets skipped constantly under pressure to ‘modernise’. AI should be the answer to a question you've already identified, not a solution in search of a problem!


I – Incremental Adoption


Don't go all guns blazing. A phased approach lets you validate whether a solution performs as promised in your environment, with your people and your data – not in a supplier’s controlled demo.


It also gives you space to manage the change properly, identify integration issues early, and build internal confidence before you're too far committed to reverse course.


Security operations carry real


consequences. The stakes are too high for an all-out approach.


L – Lift vs. Load


Every AI solution will promise efficiency gains, such as faster detection, reduced manual effort, better resource utilisation. Hold those claims up against the full cost of adoption. Not just the initial price, but the load that comes with it: staff training, integration work, ongoing maintenance, and the risk that it makes existing workflows and practices more complicated rather than simpler.


If the operational lift doesn't clearly outweigh the load, the business case simply isn't there.


T – Test for True Value


This is where you cut through the noise. Does the solution deliver genuine, measurable operational impact, or does it just make for a good capabilities slide in a supplier’s presentation deck? Push suppliers on evidence. Ask for case studies from organisations like yours. Ask what the solution can't do as much as what it can.


11 © CITY SECURITY MAGAZINE – SUMMER 2026


Meaningful innovation improves outcomes for the people delivering security on the ground. If the primary beneficiary is a sales pitch, it's not meaningful innovation.


E - Empower the Human


AI should augment human capability, not replace it. The skills, judgement, and contextual awareness that experienced security professionals bring cannot easily be replicated by a model; and in security, the consequences of removing that human layer at the wrong point can be significant.


Whether a human should be on the loop (monitoring and reviewing) or in the loop (actively involved in real-time decisions) depends on the use case and the risk tolerance involved. Either way, the human stays in the picture. Any supplier telling you otherwise rightly deserves scrutiny.


R – Right-Size the Solution


Not every problem needs a bespoke, enterprise-level AI platform. Look for where the quick wins are – applications with existing precedent, proven track records, and lower implementation risk. Consider whether a commercial off-the- shelf (COTS) tool or even open-source options could achieve what you need without the overhead of a specialist procurement.


Start where confidence is highest, demonstrate value, then build from there. Momentum matters, especially when you're bringing a wider team or senior leadership along on the journey.


One thing's for certain – AI isn’t going anywhere and the security industry will be a better place for it, eventually! Apply the FILTER, ask the hard questions, and remember: the organisations that will benefit most won't be the ones who adopted earliest. They'll be the ones who adopted wisely.


Ash Risby MSyI


Chair, AI & New Emerging Security Technologies SIG, Security Institute


www.security-institute.org www.citysecuritymagazine.com


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36