Back Issues/Search Home → Calendar → Archive → RSS → Subscribe → Services → Vendors → Blog → Current Issue → Popular → Newsletters (LinkedIn, Substack)

All issues › Volume 344, Issue 2 › IT Vendor News › Snowflake

The Hidden Cost of Moving Data for AI

Snowflake, Wednesday, November 11th, 2026

Snowflake argues that moving data before starting AI adds cost and risk, and that open formats let AI run on data where it sits.

Snowflake says CIOs are often told to migrate and centralize data before starting AI, but every move adds cost and risk, and that sequence only made sense when querying data required each system's proprietary compute.

Open table formats such as Apache Iceberg separate storage from compute, so AI workloads can run on data where it already lives while modernization continues separately.

Snowflake Horizon Catalog applies one set of governance policies across data and assets, now extends row access and masking policies to engines such as Spark, Trino and PyIceberg in public preview, and supports separate agent identities.

Snowflake notes this matters for GDPR, data residency and the EU AI Act.

more →  ·  More from Snowflake →