Artificial intelligence is moving rapidly beyond centralised cloud environments into the real world. Manufacturers are deploying AI on production lines, transport operators are embedding intelligence across distributed networks and public services are increasingly relying on AI-powered systems to automate the services they provide.
From faster decision-making to lower latency, this is offering significant benefits for UK organisations. Yet it also marks a fundamental shift in the way infrastructure is secured. Every intelligent edge device is no longer just another endpoint; it is an operational system capable of making decisions, processing sensitive data and influencing physical environments.
Yet the majority of organisations are unprepared for this shift, with less than a quarter (24%) of UK businesses currently using, adopting or considering AI having security processes in place to manage the associated risks. As edge AI expands, so does the attack surface. Organisations need to build security into their infrastructure from the onset, combining greater visibility, secure device management and faster detection and response.
Edge AI is changing what infrastructure means
For years, enterprise security strategies have focused on protecting well-defined environments, including corporate networks, data centres and managed cloud platforms. Those boundaries are becoming less relevant as AI workloads move closer to the edge.
An intelligent camera monitoring transport infrastructure, an AI-enabled robot on a manufacturing line or a medical device analysing patient data all operate in environments where connectivity may be intermittent, decisions happen locally and downtime can have immediate operational consequences.
Unlike traditional endpoints, these devices are not simply collecting information; they are interpreting it and acting on it. The challenge is not only protecting individual devices, it’s maintaining trust across thousands of distributed systems running different operating systems, software and AI models. For organisations operating critical infrastructure, edge systems need to be treated as part of core infrastructure, rather than separate hardware sitting beyond the wider security strategy.
The challenge is that most organisations still manage these environments through separate operational functions. Security teams focus on threats, operations teams focus on availability and cloud teams focus on performance. At the edge, those distinctions become increasingly artificial. At the edge, a security event, an infrastructure event and an application event are often the same incident viewed through different operational lenses. Responding effectively requires shared visibility and coordinated decision-making across all of them.
Security and operations can no longer work separately
As the attack surface expands, the threat tactics used by attackers are evolving too. Cybercriminals are increasingly using automation and AI to accelerate reconnaissance, identify vulnerabilities and scan environments at scale. The challenge for organisations is therefore not just identifying threats but doing so before they can spread across the wider network.
IT teams are also often overstretched, with network monitoring, security detection, configuration management and incident response operating through separate systems. While each may address a specific problem, this fragmentation can make it harder to identify whether an alert is isolated or part of a wider incident.
AI-driven detection and response can analyse activity across an organisation’s infrastructure, identify patterns that may be missed manually and prioritise the incidents that matter most. This gives teams the context to respond more quickly without working through every signal themselves. More importantly, it helps teams move beyond the fragmented workflows that have traditionally separated security, operations and infrastructure management. As AI-powered systems become distributed across edge environments, organisations need a shared operational understanding of risk, performance and availability rather than separate views maintained by individual teams. The aim is not to replace human expertise, but to augment it, helping experienced teams focus on genuine threats while shifting toward more proactive infrastructure management.
Infrastructure management provides the foundation
Detection alone will not secure edge AI; organisations need confidence that devices can be securely managed once deployed.
Trusted communication between edge devices and central management systems is essential. Organisations need reliable ways to deploy updates, enforce policies and maintain visibility without unnecessarily expanding remote access. Approaches that maintain a secure, persistent management channel can also help support continuity when connectivity is unstable, allowing approved tasks to continue until communication is restored.
Device management also cannot become another isolated layer. Monitoring, network detection, response and configuration systems need to exchange information and contribute to a shared operational view. Buying individual products and attempting to connect them afterwards can create gaps in defence and add complexity for already overstretched teams. SecOps, Network Operations, CloudOps and DevOps also need to share intelligence and coordinate decisions across the same AI-enabled infrastructure, rather than operating through different workflows.
Edge AI will increasingly underpin the UK’s digital and physical infrastructure, making secure management a core operational requirement rather than an afterthought. Success will depend less on deploying individual security products and more on connecting operational intelligence across the environment. Security telemetry, network visibility, device management and infrastructure monitoring all need to contribute to the same picture if organisations are going to manage AI-powered systems at scale.
Organisations that succeed will be those that reduce fragmentation, maintain control across distributed environments and give teams the tools to act with greater speed and confidence. By treating AI-enabled devices as part of wider infrastructure, organisations can scale adoption without compromising on security.
Frank Cotto
Frank Cotto is Field CTO, Infrastructure Management and Operations at Progress Software where he is focused on building trusted relationships and helping organisations observe, automate, secure and scale modern hybrid infrastructure using AI-driven insights.
