Databricks just flipped the script on data and AI collaboration — and the ripple effects will be felt across every enterprise that’s ever been trapped behind a vendor’s walled garden.
On June 11, 2026, Databricks announced OpenSharing — the next evolution of its widely adopted Delta Sharing protocol, now hosted by the Linux Foundation and expanded to cover the full AI stack.
Table of Contents
What Is OpenSharing, Exactly?
Think of OpenSharing as a universal plug for data and AI assets. Where Delta Sharing broke silos for structured data, OpenSharing goes further — covering Agent Skills, AI models, and unstructured data through one open, vendor-neutral protocol.
“The agentic era deserves an open foundation, and OpenSharing delivers it.” — Matei Zaharia, Co-founder & CTO, Databricks
The project is live on GitHub and open to the world.
What OpenSharing Brings to the Table?
| Feature | What It Means for You |
|---|---|
| 🤖 AI Asset Sharing | Share agent skills & AI models across orgs — no custom integrations |
| 🌐 Publish Once, Reach All | Works with Delta Sharing AND Apache Iceberg™ clients |
| 🏢 On-Premises Support | Connect on-prem data to cloud AI — zero data movement |
| 🔐 Zero-Copy Security | Data never leaves the source; access is authorized, not duplicated |
| 🔓 Vendor Neutral | Hosted by Linux Foundation; no single company controls it |
Why This Is a Big Deal in Databricks ?
Before OpenSharing, enterprises had two painful choices: build expensive custom integrations to share AI assets across platforms, or lock themselves into a single vendor’s marketplace.
OpenSharing kills that dilemma.
With on-premises storage partners like Everpure, MinIO, and Qumulo already onboard — and Cohesity, HPE, NetApp, Nutanix, and Rubrik coming soon — companies no longer have to choose between on-prem security and cloud-scale AI.
Who’s Already Using It?
The industry response has been anything but quiet. Major players are already aligned:
- Amadeus — Orchestrating trusted AI across the global travel ecosystem
- Atlassian — Powering Analytics with seamless data shares from its Data Lake
- LSEG — Letting customers use their data in any tool, any cloud, any model
- SAP — Baking OpenSharing into SAP Business Data Cloud
- Stripe — Unlocking secure analytics on billing and transaction data
- OpenAI — Collaborating on standard discovery and authorization for AI assets
This isn’t a beta rollout — it’s a full-scale industry shift
How It Connects to the Larger AI Ecosystem
OpenSharing extends Delta Sharing’s already-massive cross-platform reach. Today, Delta Sharing users can receive shares in Apache Spark, Oracle, Power BI, Tableau, Snowflake, and more. OpenSharing now adds Apache Iceberg™ API support, widening the net further.
For a deeper dive into how Databricks is shaping the AI infrastructure landscape, check out our coverage of Databricks’ latest AI platform updates on TechnoSports.
FAQs
Q: Is OpenSharing free to use?
It’s open source and hosted by the Linux Foundation, making it freely available for enterprises and developers to adopt without licensing fees.
Q: How is OpenSharing different from Delta Sharing?
Delta Sharing handled structured data; OpenSharing extends that to AI models, agent skills, unstructured data, and now Iceberg-native clients too.
