POV: Sovereignty-Driven Lakehouse Demand in Europe – Critical Entry Window?

With EU AI Act enforcement beginning August 2026, European clients are no longer just “concerned” about sovereignty — they are legally liable for where models run and how data is governed. This feels less like a small window and more like a critical entry window for IBM.

We’re seeing a widening “SaaS gap”:

• Databricks and Snowflake are positioning sovereign cloud wrappers — often contractual sovereignty (legal assurances).
• IBM can offer architectural sovereignty — where clients retain technical control over identity, encryption keys, and audit logs within their own jurisdiction.

Specifically:

• Sovereignty by Design, not by Proxy.
• Clients retain authority over identity, encryption keys, and audit logs within their own jurisdiction.
• No US-based control plane dependency.

watsonx.data becomes highly relevant here:

• Built on open formats (Iceberg, Presto) → no proprietary data lock-in.
• Deployable self-managed in-region or on-prem.

When combined with:

• IBM Power10 (confidential computing + memory-level encryption)
• Storage Fusion (air-gapped, on-site data control)

We can offer true Sovereign Operations, not just Sovereign Regions.

This aligns strongly with IBM’s broader Sovereign Core proposition — but the Lakehouse layer feels like a near-term wedge for our teams.

Are we tracking sovereignty-driven self-managed lakehouse demand as a strategic signal in EMEA?
And do we have a formalized “Sovereign AI-Ready Data Platform” narrative before this enforcement cycle accelerates decisions?