The vast AI investments of enterprise firms are starting to pay off, but CIOs are struggling to scale those gains into organisation‑wide transformation, and are unprepared for the next wave of disruption, according to fresh research from Lenovo.

Polling over 3,000 IT and business decision makers around the world, Lenovo’s Race for Enterprise AI found that almost half (46%) of AI proofs-of-concept have progressed to production, with some firms anticipating returns of $2.79 for every dollar invested.

Strong returns may have cemented AI as a central driver of business reinvention, but the research flags a clear overconfidence among Chief Information Officers (CIOs), whose readiness limits progress and could undermine future plans.

Lenovo found that while 60% of organisations are in late-stage AI adoption, only a quarter (27%) have developed a robust governance framework, a problem compounded by a lack of in-house expertise, integration complexity, and weak organisational alignment over AI strategy.

Despite those fundamental problems, however, CIOs anticipate up to 179% ROI on current AI investments and predict more efficiencies as agentic AI spending overtakes generative AI. 

Nearly all (96%) of those surveyed plan to increase their AI investments in the next twelve months to focus on such next-gen systems, with 93% anticipating positive returns.

However, that conviction is unfounded, according to Lenovo’s survey, with three in five (60%) of CIOs admitting they are more than a year away from being ready to scale agentic AI across their operations. 

So far, only 21% report significant use of agentic AI systems, while the majority (55%) are still piloting or actively exploring use cases.

“As AI priorities shift toward agentic AI, the next phase will not reward experimentation – it will reward those able to operationalise AI across hybrid environments with trust and scale built in,” said Ken Wong, president of the solutions & services group at Lenovo.

Recommended reading

• 69% of C-Suite Leaders Choose Fast AI Over Secure AI
• 80% of Firms Say Their AI Agents Have Taken Rogue Actions
• Capgemini: AI Adoption Shifting Focus to Governance & Leadership

The study also found increasing interest in hybrid AI, an operating model that blends public cloud, private cloud, and on-premises compute, with 84% of firms now using on‑premises or edge deployments for AI workloads as part of a broader hybrid environment.

With AI PCs and edge endpoints central to an effective hybrid AI strategy and securely running AI workloads locally, deploying AI-capable devices has emerged as the top IT investment priority for 2026.

“This year’s research makes it clear that AI devices and edge endpoints are now the frontline of enterprise AI in powering every employee, securing workflows, and putting intelligence exactly where work happens,” said Luca Rossi, Lenovo’s president of intelligent devices group.

“When combined with the right infrastructure and services, this end-to-end approach gives enterprises what they’ve been asking for: a way to innovate confidently, responsibly, and at scale.”