Why AI and cost pressure make multi-cloud interoperability critical (www.techradar.com)

Exploding data volumes and rising AI experimentation are stretching IT budgets and exposing weaknesses in single-vendor cloud strategies. Organizations face surging storage needs (global data rising from 149ZB to 181ZB), higher cloud spend, and escalating cyber and geopolitical risks — from ransomware outages to data sovereignty concerns that make hosting with non-local hyperscalers legally and operationally risky. At the same time, AI workloads push requirements for expensive GPUs, greater power and cooling, and high‑spec infrastructure, driving up costs and locking teams into proprietary stacks that reduce visibility and mobility. The answer gaining traction is multi‑cloud interoperability: running the right workload on the right cloud to optimize cost, resilience and compliance. Nearly 80% of UK firms are evaluating or using multi‑cloud, citing flexibility and cost savings, and practical wins (e.g., Hopsworks cutting cloud spend by 62%) show it can work. But multi‑cloud is complex and requires open standards, transparent pricing and tooling to avoid vendor lock‑in — hence the importance of groups like OpenUK, Open Cloud Coalition and OpenStack and regulatory scrutiny from bodies like the CMA. For the AI/ML community, embracing open, portable infrastructure is essential to control costs, preserve competition, secure critical systems and sustain scalable, auditable AI deployments.