On 18/06/2026 at the Databricks Data + AI Summit, xAI announced that Grok is now natively available on Databricks (Agent Bricks). The notable part: AI agents can reason directly on data residing in the enterprise Lakehouse — both structured and unstructured data — without having to push it through an external pipeline.
Quick summary
- When: 18/06/2026, at the Databricks Data + AI Summit.
- What: Grok (xAI) is now natively available on Databricks.
- Meaning: agents reason directly on data inside the Lakehouse, without pushing it outside.
- Trend: 'bring AI to the data' instead of 'bring data to the AI'.
What happened?
According to xAI's announcement and reports from the event, Grok is now one of the models you can select right inside the Databricks environment. Instead of copying data out to an external AI service and then processing it, an enterprise can let an agent run right next to the data within its own governed platform.
Why this matters
This is a clear expression of the "bring AI to the data" trend rather than the other way around. When an agent reasons right inside the governed data zone (with the Lakehouse's built-in governance, access control and auditing), the enterprise reduces the risk of data exposure during movement and retains better control.
A perspective for enterprises
This spirit aligns with Namtech's positioning: AI works on data that stays within your walls. With a private internal AI platform, an AI assistant + RAG runs directly on your internal documents and data, never sent outside — a fit for enterprises that value data sovereignty.
Frequently asked questions
What does Grok on Databricks mean?
It means Databricks users can choose to use xAI's Grok model right inside the platform, so agents can reason on data in the Lakehouse without having to push that data out to an external service.
How does this relate to data sovereignty?
When AI runs next to the data inside a governed zone, the data has to move outside less, which improves control and compliance.
Can Vietnamese enterprises adopt this?
It depends on the infrastructure. For enterprises that want to keep all of their data on-site, an on-premise internal AI model is an option to achieve a similar goal.
Bring AI to your data
Namtech deploys an AI assistant + RAG that runs right on your internal data — without pushing data outside.
Book a free consultationNote: This article is compiled from public sources as of 23/06/2026; information is for reference and may change.