Five Open Table Trends You Can’t Ignore in 2027

For most of the last decade, the data lake had one job: hold everything, cheaply, until someone needed it. Apache Iceberg and Delta Lake have made that job respectable, bringing transactions and schema evolution to files on object storage.

What comes next is pressure. Operational databases now write directly into open tables. Streaming platforms materialize topics as tables with no pipeline in between. And AI agents now query those tables in loops, at machine speed.

We reviewed spec votes, release notes, vendor documentation, analyst predictions and surveys published through fall 2026. Five trends stand out. Together they tell a three-act story about a lake being asked to do jobs it was never designed for.