Skip to content

Companies Shift Productive AI Workloads Out of Public Cloud

Share on:

In brief: 56 percent of companies operate or plan productive AI inference in private cloud, while public cloud usage declined by 15 percentage points globally; Germany saw a more pronounced drop of 24 percentage points.

The Broadcom Private Cloud Outlook 2026 report documents a trend reversal: companies are shifting AI inference workloads out of public cloud environments and increasingly operating them in private infrastructures – driven by cost savings and data sovereignty.

Current data from the Broadcom report shows a clear shift in productive AI deployment: 56 percent of surveyed companies already operate their AI inference in a private cloud or plan to do so. In contrast, public cloud usage for productive AI workloads declined globally by 15 percentage points from 56 to 41 percent. In Germany, this trend is more pronounced: public cloud usage for productive AI inference fell from over 56 to 32 percent – a decline of 24 percentage points year-over-year.

The migration back is primarily driven by economic factors. 99 percent of German IT leaders expect that a portion of their public cloud spending goes unused. 47 percent of these companies estimate the financial waste at over 25 percent of their entire public cloud budget. As a result, 81 percent of German companies are considering shifting workloads back to private cloud, while 44 percent have already completed this process. At the same time, planned investments in private cloud infrastructure have grown twice as fast as spending on public cloud solutions. For 61 percent of German IT decision-makers, building new workloads in private cloud is a top priority.

Beyond cost considerations, geopolitical factors are gaining influence. Four out of five IT leaders surveyed globally report that geopolitical developments directly influence their IT strategy. In Germany, data sovereignty and data residency requirements now take priority with 59 percent, ahead of country-specific compliance at 52 percent as the most important geopolitical factor. This trend is particularly driven by heavily regulated sectors such as financial services, public administration and healthcare, which must maintain control over sensitive data assets when scaling AI.


Source: www.it-daily.net · Published June 18, 2026
Lumi AI News — AI-assisted curation in accordance with Article 50 EU AI Act. Paraphrasing and classification by Lumi News Pipeline v1.7.1.

Share on: