This article is part of our Opinions section.
2025 was the year AI moved from hype to reality. From chatbots in our day-to-day to enterprise-scale deployments, most organisations experimented with AI in some way. As we look ahead to 2026, I believe one element will always remain true. AI succeeds only when the foundations below it are strong. This article shares the key fundamentals I believe IT leaders should prioritise nailing these fundamentals, which apply to every transformational project AI or not!
Back to fundamentals
AI doesn’t run in isolation – it runs on your network. The words of Cisco CTO Chintan Patel could not be truer: “there is no AI without the network”. If your infrastructure isn’t ready, every initiative will stumble. Before deploying tools like Microsoft Copilot or custom AI agents, ask:
- Can your network handle the surge in traffic from anywhere?
- Is it secure, global and optimised for performance, including voice and real-time apps?
- How standard is the infrastructure? Are they different solutions or setups in each office or region?
AI‑ready networks: what good looks like
- End‑to‑end performance: size, bandwidth and backhaul for bursty AI traffic; apply QoS so AI doesn’t starve voice and critical apps.
- Wireless‑first done right: design for density, roaming and latency; validate with site surveys and continuous monitoring.
- Secure by default: adopt zero‑trust principles; enforce strong identity, encryption and segmentation across office and remote users.
- Global connectivity: leverage SD‑WAN/SASE to deliver consistent policy and performance across regions.
- Observability: baseline now, then instrument the network (telemetry, flow data) to spot AI traffic patterns early and tune accordingly.
Don’t forget security
Security cannot be an afterthought. Make connectivity seamless for users while ensuring traffic is protected and encrypted. Policies must follow the users and contractors from the office, home, or on the move.
While also ensuring that it is possible to observe and adapt quickly, XDR or SIEM solutions should be considered as part of the overall strategies.
Operational support: empower teams without overload
Cutting‑edge initiatives like AI can consume IT operations time. Your internal teams carry critical in-house knowledge, so involve them early to avoid delays and rework. Where capacity is tight, co‑managed services can help without the downsides of fully outsourcing. The goal is a trusted partner who works shoulder to shoulder with your team on proactive operations.
- Focus where it matters: free internal teams from patching and routine break/fix so they can drive transformation.
- Integrate processes: 3rd line to 3rd line collaboration, shared ticketing and joint SLAs to operate as one team.
- Proactive posture: detect and remediate issues before users feel them via proactive monitoring, automation and runbooks.
- Time to value: AI projects with on‑prem LLMs can be hardware‑intensive and costly, with surges in pricing from demand, such as the impact for RAM being felt right now. Speed of rollout drives ROI quicker!
Cost optimisation: fund innovation by simplifying
Doing more with less isn’t new, but it’s essential for 2026. The fastest way to fund new initiatives like AI is to remove waste and negotiate smarter. Start with a clear map of your current estate and spending, then consolidate what overlaps.
- Contract and spend reviews: look across maintenance, cloud bills and renewals over multi‑year horizons to find savings.
- Tool consolidation: avoid duplicated capabilities (eg endpoint protection included in platform licences plus a separate product will result in overlapping tooling).
- Smarter buying: Use enterprise agreements and committed use discounts to secure better pricing for the tools you truly need.
- Cross‑functional alignment: bring IT and procurement together. They are working on the same goals, but may feel they are against each other. The value to the business must be shared with them, working as one team.
When you unlock savings, you can reallocate budget to the initiatives that move the needle, like responsible AI rollouts.
2026 — The year of fundamentals
Expect AI to keep dominating headlines – but success will hinge on fundamentals. Here’s the concise checklist:
- AI‑Ready Networks: secure, performant, wireless‑first and globally consistent.
- Proactive Managed Services: trusted partners who integrate with your team to accelerate change.
- Cost Savings: reduce operational and tooling costs to fund innovation, including AI.
I’m excited to see how we use technology in 2026 to transform the IT experience. Grounded in strong fundamentals and ready for what’s next.
Read our verdicts on 2025 by worldwide region
Trending
