Global IT expenditures on generative artificial intelligence are undergoing a structural shift. According to recent enterprise survey data, top-tier financial and healthcare firms are shifting up to 35% of their AI workloads back to on-premises data centers.
Cost Realities and Regulatory Pressure
The primary drivers behind this migration are unpredictable cloud API costs and stringent data sovereignty laws in the European Union and Asia-Pacific regions. While cloud hyperscalers offer rapid deployment, high-volume inference costs are proving unsustainable for large enterprise balance sheets over long time horizons.