Stephen McGowan, Regional Leader UKI and South Africa at Progress Software, offers three ways infrastructure leaders can build cyber resilience for AI.
As organisations scale AI, enterprise IT infrastructure will become increasingly business-critical over the next twelve months. Infrastructure leaders are facing three interconnected pressures as a result: maintaining visibility and security across complex environments, ensuring infrastructure can support AI at scale, and managing the associated costs. These priorities cannot be addressed in isolation.
To be truly AI-ready, organisations need infrastructure that is fully observable and secure with cyber resilience built in from the outset. The challenge for leaders, then, is to modernise at pace so their infrastructure can keep up with the demands of AI.
Fragmented infrastructure creates unexpected blind spots
Organisations are operating their IT infrastructure across an increasingly fragmented mix of systems, from legacy and on-premises environments to multicloud platforms. As applications and data become more dispersed, leaders must ensure that expanding systems do not come at the expense of visibility and control. Legacy infrastructure will continue to coexist withemerging technologies like AI, making it increasingly important for organisations to understand how their systems interact and where blind spots may emerge.
The rise of agentic AI makes this visibility even more critical. As AI agents gain greater autonomy to act across systems, IT teams need clear, measurable processes to monitor their activity, permissions and access privileges. Without appropriate controls, agents could be given excessive permissions or take unintended automated actions. Organisations therefore need clear audit trails showing what agents have done, which systems and data they have accessed and where human oversight is required.
Resilience also means being able to intervene when something goes wrong. Autonomous systems need mechanisms that allow teams to contain, stop or reverse an agentโs actions where necessary. This is particularly important for organisations handling sensitive data as regulation around AI and cybersecurity continues to evolve.
Developments such as the UK Cyber Security and Resilience Bill reinforce the importance of adopting a secure-by-design approach. Yet many organisations are still developing the processes needed to manage AI-related cyber risks. In a survey published by the UK Government in April 2026, only 24 per cent of businesses already using, adopting or actively considering AI reported having cybersecurity processes in place. As adoption accelerates, organisations will need to build security, incident response and oversight into AI deployments from the beginning. Maintaining visibility, control and reliability across both AI systems and the infrastructure they depend on will be critical as IT environments become increasingly autonomous.
AI ambition overshadows infrastructure readiness
Recently, the Confederation of British Industry called on the UK Government to make AI adoption an economic priority. For leaders, this poses a significant challenge: equipping their IT infrastructure to keep pace with their organisationโs AI ambitions.
As organisations move beyond AI experimentation and embed the technology into day-to-day workflows, the demands placed on their infrastructure will grow. AI at scale requires greater compute capacity, reliable access to data and consistent performance across complex environments. Without the right foundations in place, ambitions to scale AI could quickly run into operational and performance constraints.
As these demands grow, leaders should consider how AIOps and autonomous operations can help them proactively monitor performance and manage increasingly complex environments. Infrastructure readiness must be treated as part of AI adoption rather than an afterthought, with systems designed to remain observable and governable as they scale. As pressure from boards and investors to manage AI spend increases, organisations that scale too quickly without adequate foundations risk creating a disconnect between their AI ambitions and their infrastructureโs ability to support them.
The cost of cutbacks
Over the next year, leaders will face mounting pressure to cut costs and demonstrate a clear return on investment (ROI). However, pursuing short-term savings risks leaving organisations overstretched and without the resources needed to manage increasingly complex IT environments. Recent research from McKinsey suggests that one in five companies are limiting AI use due to operating costs. With AI spending likely to face greater scrutiny from investors and boards, leaders will need to be more strategic about which projects they pursue and consider whether short-term savings could create greater operational costs over time.
One risk is cutting headcount without retaining the skills and expertise needed to govern increasingly automated IT systems. AI and automation can reduce the need for some manual tasks, but human oversight remains critical to understanding the decisions and actions these systems take. Strong governance, monitoring and mechanisms for human intervention become particularly important when systems fail or behave unexpectedly. Likewise, leaders may face pressure to choose enterprise infrastructure options that appear more cost-effective in the short term. But decisions made primarily on upfront cost can introduce greater complexity, create gaps in stability and increase operational costs over time.
Ultimately, infrastructure maturity will become increasingly important as leaders balance technology investment with financial pressures and wider business goals. Organisations that can demonstrate ROI without losing sight of the resilience of their infrastructure will be better placed to meet board and investor expectations while supporting AI at scale.
Looking to the future
These three challenges are closely connected: automation can make it harder to maintain visibility, scaling AI places greater demands on infrastructure, and cost-cutting can weaken reliability and stretch resources. Leaders therefore need to take a joined-up approach, maintaining visibility across their environments, ensuring infrastructure is ready for growing AI demands and making investment decisions with resilience in mind. By addressing these in tandem, organisations will be better placed to balance AI ambition with infrastructure readiness and scale the technology safely in the long-term.
