AI Ops Console Sprawl 2026: Gartner's Warning to UK IT

What Gartner's Hype Cycle actually predicts

Gartner published its 2026 Hype Cycle for AI in IT Operations on 10 July, setting out how AI-powered infrastructure management tools will evolve through the decade. The headline figure is stark for planning purposes: a quarter of all work currently done by IT infrastructure and operations staff will be handled by AI by 2030.

But the analyst firm is explicit that the well-worn pitch of "AI will consolidate your tools" doesn't hold up in the near term. Gartner's own document states that agents are supposed to "query multiple systems, reason across silos and reduce dependence on specialized tools" — before predicting "the opposite outcome in the near term." For UK buyers building multi-year roadmaps, that's a direct instruction to stop budgeting for consolidation savings this year or next.

More consoles, not fewer, in the near term

Rather than shrinking your toolset, Gartner expects organisations to accumulate "more layers, more control points, and more specialized observability, orchestration and management capabilities" for at least the next couple of years. Only later, once market and vendor consolidation kicks in, does the tooling footprint shrink.

That's a costly interim phase for any UK estate already juggling multiple vendors, contracts and support agreements. Teams that already manage a multi-vendor IT estate with third-party maintenance should assume the AI ops layer adds new consoles on top, not fewer — and plan procurement and support renewals accordingly rather than assuming near-term simplification.

The outage numbers UK ops leaders can't ignore

Gartner's most consequential figure for risk planning: by 2028, 40 percent of infrastructure and operations organisations running agentic AI at scale in production will suffer a business-critical service disruption — up from under 1 percent in 2026. That's a jump analysts want written into strategic planning assumptions now, not discovered after the fact.

It won't slow deployment. Gartner expects 60 percent of enterprises to run agentic AI as part of IT infrastructure operations by 2029, up from fewer than 10 percent today, while human sign-off on AI-suggested actions falls from 80 percent in 2025 to just 20 percent by 2029, as organisations lean on policy-driven "deterministic guardrails" instead of manual approval. For boards, this reframes outage risk as a budget line rather than an edge case — worth running through a calculate the potential cost of downtime exercise before signing off wider agent deployment, and worth reading alongside work to understand the real cost of IT downtime in a UK context.

Which AI ops technologies mature first — and which matter most

Gartner rates several technologies as likely to mature within the next couple of years, giving buyers a shortlist of what's realistically procurable soon rather than speculative.

Looking further out, Gartner names four technologies reaching maturity in two to five years that it expects to be the most impactful: Agentic AI Observability, Agentic NetOps, Augmented FinOps and Multiagent Systems. Agentic AI Observability — tooling that watches AI agents and flags when they misbehave — stands out as the one to prioritise first, since it underpins both cost control and governance for everything else on this list.

• •GenAI-native IT ops vendors that build tools around generative AI from the ground up, rather than bolting it onto legacy platforms
• •GenAI virtual assistants offering conversational self-service that connects to agents to trigger fixes
• •GenAI-augmented CloudOps analysing logs, metrics, traces and change events to auto-generate scripts, infrastructure-as-code, runbooks and incident reports
• •Autonomous endpoint management that configures machines to user profiles and keeps pace with rising patch volumes
• •Network AI and automation tools that recommend configuration changes and offer conversational interfaces for networking kit

Governance before you buy: the two-tier model

Gartner's guidance leans hard on governance as the fix for sprawl, recommending a two-tier structure: a centralised AI governance committee — CIO, CISO, chief AI officer, architects, legal and business leaders — sets strategy, while domain-specific operational teams enforce controls within each application area.

Underneath that sits practical hygiene most UK estates still lack: a centralised inventory of every AI agent, sanctioned and shadow, built using AI TRiSM (AI Trust, Risk and Security Management) tooling; a defined identity and least-privilege permissions for each agent; and a lifecycle plan that retires redundant agents rather than letting them accumulate. That last point matters — roughly 40 percent of organisations are already planning to demote or decommission AI agents because governance hasn't kept pace with deployment. Gartner has separately warned that misconfigured AI embedded in national infrastructure could shut down critical services in a G20 nation by 2028, disruption it compares to a hostile cyberattack or natural disaster. Performance data adds weight to the caution: IBM's own open-source agent, CUGA, currently manages only a 61.7 percent success rate on web tasks and 48.2 percent on API tasks — a level IBM itself describes as poor enough to get a human worker fired.

What UK infrastructure buyers should do now

None of this argues against agentic AI adoption — Gartner's numbers show it's coming regardless. It argues for sequencing governance ahead of scale, and for treating vendor consolidation promises as a multi-year outcome rather than a launch-day feature.

Before extending agent permissions into production, map the customer or operator journey you're automating and set guardrails at each decision point, rather than granting broad autonomy up front. Teams reviewing their wider estate strategy can discover comprehensive IT solutions built around this staged approach, and should explore more AI infrastructure insights as Gartner's maturity timelines shift.

• •Build a full inventory of sanctioned and shadow AI agents before adding more
• •Classify agents by capability and risk profile to match governance rigour to each tier
• •Prioritise agentic AI observability tooling before expanding agent scope
• •Treat the 2026–2028 window as a sprawl period to budget for, not avoid
• •Revisit third-party maintenance and support contracts to cover new AI ops layers