What the survey is
Dell's Modern Enterprise Readiness Study is a Dell Technologies and Vanson Bourne survey of 2,950 business and IT decision makers across 35 countries. The research was conducted in June 2026. The EMEA portion covered 1,500 decision-makers.
TechRadar's report of the study singles out UK firms. It says 61% of British firms lack a clear roadmap spanning AI, data and security, against 46% across EMEA. I couldn't confirm those regional figures in Dell's public material. Treat them as TechRadar's account of the Dell data.
Dell's global figure is lower. Its global material says 66% of respondents believe their infrastructure isn't fully ready for demanding AI workloads at production scale. TechRadar reports 64% of IT leaders saying the same. The gap probably comes from a different subgroup or different wording, so don't treat the two numbers as the same measurement.
Compute is not the main bottleneck
The most useful finding for admins is that GPUs and servers aren't the sticking point for many respondents. Dell reports that 80% say storage performance and data access are now bigger bottlenecks than raw compute for critical workloads.
That is a survey opinion, not a measured ranking. It still matches what many IT teams report in practice, in three ways:
- Data is scattered across file shares, SaaS apps and legacy databases.
- Permissions and classification are inconsistent.
- The pipeline from storage to workload is slower than the model that needs the data.
Dell's own write-up makes the same point. Its full findings say storage performance now outranks raw compute as the bigger bottleneck for many critical workloads.
Security and compliance are slowing projects
The survey also points to caution about regulation and data exposure. The Dell global figures and the TechRadar figures differ here too:
| Measure | Dell global figure | TechRadar report |
|---|---|---|
| Security/compliance slowed or paused AI | 60% | 55% |
| Worried about regulatory compliance when using data for AI | 68% | 64% |
Dell's global material adds a few more numbers. 68% say their data and intellectual property are too valuable to put into a generative AI tool where a third party might access them. 74% would only scale AI without significantly increasing their energy footprint. 82% call energy efficiency essential when buying infrastructure.
These are reasons to slow down that many IT leaders will recognize. Regulators, auditors and legal teams rarely accept "we'll fix governance later". Dell's lesson is to build security into the design from the start. That is also the message of a vendor selling security-rich infrastructure, so weigh it accordingly.
Where AI PCs fit
TechRadar says 92% of respondents agree modern devices help productivity, and 79% say outdated devices are holding them back. It also says 86% of EMEA leaders expect AI PCs to matter to their strategies. I couldn't verify those exact figures in Dell's public pages.
The full Dell findings do give a picture of fleet makeup. The average fleet is about 22.6% older devices (4+ years), 36.6% modern but not AI-capable devices, and 40.8% modern, AI-capable devices. That makes the typical respondent's fleet about 41% AI-capable. The finding does not mean 41% of those PCs are running AI workloads.
TechRadar also says AI PCs were about 44% of PC shipments at the start of 2026, citing Omdia. I couldn't tie that number to a global or EMEA Omdia source. The 44% Omdia figure I found related to AI-capable notebooks in India in Q2 2026. Read the shipment claim cautiously.
An AI PC is useful for specific jobs. Dell's own pitch is that local hardware can run some AI tasks without the cloud. That can help with latency, offline use and keeping some data on the device. It does nothing for the problems the survey keeps raising: data quality, governance, storage performance and compliance. A PC refresh is not a substitute for fixing those.
What IT teams can take from it
Dell's own suggested starting points are reasonable, even if they come from a vendor:
- Assess the data foundation, storage performance and security posture before scaling.
- Treat security as a design requirement, not a final checklist.
- Ask partners to show business impact and a defined risk-management plan.
A few caveats apply. Survey respondents are decision-makers, and their answers reflect perception. The study also leans toward consolidating on fewer, integrated vendor platforms, and Dell sells those platforms. Treat the vendor-consolidation argument as Dell's interest as well as a finding.
The takeaway
The most credible message in the data is that AI readiness is a pipeline problem. It runs from data location and access, through storage and security controls, to compliance and energy limits, and only then to the endpoint. Organizations that skip the early steps and buy AI-branded hardware first are likely to find that the survey's bottlenecks are still there.