U.K., London, November 16, 2026
AI's Elusive ROI: How to Track, Manage and Demonstrate Value
December 17 2026
U.K., London, November 16, 2026
It’s not too late to join the conference
Overview
We are bringing you news and highlights from the Gartner IT Infrastructure, Operations & Cloud Strategies Conference, taking place this week in London, U.K. Below is a collection of the key announcements and insights coming out of the conference.
On Day 1 of the conference, we are sharing actions to help infrastructure and operations (I&O) leaders to scale agentic AI and enable autonomous IT operations, exploring five scenarios I&O leaders should plan for 2027, and discussing the dangers of sovereignty. Be sure to check this page throughout the day for updates.
Key Announcements
Presented by Erin Pierre, Senior Principal Analyst, Gartner and Joe Rogus, Director Analyst, Gartner
Demand has shifted to scaling agentic AI beyond experiments, representing a significant opportunity to successfully deliver autonomous IT operations future. In this session, Erin Pierre, Seniorr Principal Analyst at Gartner and Joe Rogus, Director Analyst at Gartner, shared specific actions that infrastructure and operations (I&O) leaders can take to scale agentic AI.
Journalists can receive additional information and/or request an interview with Erin Pierre or Joe Rogus by contacting Laurence Goasduff at laurence.goasduff@gartner.com.
Presented by Enrique Castera, Senior Director Analyst, Gartner
AI infrastructure choices will diverge in 2027 based on adoption speed, model strategy and risk constraints. In this session, Enrique Castera, Senior Director Analyst at Gartner, explored five scenarios that change compute, data, network, security, and operations requirements.
- Scenario 1: Unpredictable token consumption makes cost forecasting impossible.
- Scenario 2: Production model serving pushes workloads into hybrid infrastructure.
- Scenario 3: On-prem AI infrastructure scaling hits kilowatts, cooling, and skills bottlenecks.
- Scenario 4: Generative AI extends to the edge for localized execution.
- Scenario 5: Production AI demands full-stack, cross-layer observability.
“If your engineering teams are blocked by run-away token costs, delayed by two-year data center retrofits, or struggling to manage scattered edge devices, your business simply cannot execute.”
Journalists can receive additional information and/or request an interview with Enrique Castera by contacting Laurence Goasduff at laurence.goasduff@gartner.com.
Presented by Alessandro Galimberti, VP Analyst, Gartner
Organizations are considering geopatriation as a way to reduce their sovereignty risk exposure. In this session, Alessandro Galimberti, VP Analyst at Gartner, discussed how to successfully balance sovereignty aspirations with the reality of a tense geopolitical environment.
“Sovereignty is a spectrum that a range of tailored solutions based on individual workload needs, not a one-size-fits-all model.”
“Strict data sovereignty fails during severe emergencies making organizations rethink sovereignty ambitions.”
“Sovereignty in open source only exists as long as big tech interests align with yours.”
“Building a fully sovereign stack is unrealistic today and unlikely in the future; for vital workloads map your dependencies and leverage commodity technologies wherever possible.”
Journalists can receive additional information and/or request an interview with Alessandro Galimberti by contacting Laurence Goasduff at laurence.goasduff@gartner.com.
It’s not too late to join the conference
Laurence Goasduff
Gartner
laurence.goasduff@gartner.com
Gartner (NYSE: IT) delivers actionable, objective business and technology insights that drive smarter decisions and stronger performance on an organization’s mission-critical priorities. To learn more, visit gartner.com.