Databricks announced OpenSharing at Data + AI Summit 2026, describing it as the next evolution of Delta Sharing for the agentic era. Hosted by the Linux Foundation, OpenSharing is the first open, vendor-neutral protocol for sharing AI assets - including agent skills, AI models, and unstructured data - in addition to the structured data sharing inherited from Delta Sharing. The protocol adds Apache Iceberg IRC client support, expanding the recipient universe beyond existing Delta Sharing connectors. On-premises and private-cloud data assets can now connect directly to cloud AI platforms without data movement, through storage partners including Everpure, MinIO, and Qumulo; Cohesity, HPE, NetApp, and others are announced for future availability. OpenAI, SAP, Stripe, Atlassian, LSEG, and Amadeus are named as backing partners at launch. The project is available on GitHub and natively within the Databricks platform.
What happened
Databricks announced OpenSharing at Data + AI Summit 2026 on June 10, 2026. Hosted by the Linux Foundation, OpenSharing is the next evolution of Delta Sharing, the open data-sharing protocol Databricks launched in 2021. Delta Sharing addressed secure, zero-copy sharing of structured tables across platforms. OpenSharing extends the open standard to cover AI-era assets: agent skills, AI models, and unstructured data, in addition to the structured sharing capabilities inherited from its predecessor.
Three additions over Delta Sharing
The announcement details three capabilities not present in Delta Sharing. First, OpenSharing provides the first open, vendor-neutral protocol for sharing AI assets across organizations. Prior to this, enterprises had no standard mechanism for sharing agent skills or AI models, forcing reliance on custom integrations or single-vendor marketplaces; OpenSharing provides a single protocol with standard APIs for discovery, authorization, and access, regardless of platform. Second, OpenSharing adds support for Apache Iceberg IRC clients, expanding the recipient universe to include Iceberg-native tooling beyond the existing Delta Sharing connector ecosystem. Third, on-premises and private-cloud data assets can now connect to cloud AI platforms without data movement, through native storage partner integrations. Partners available today include Everpure, MinIO, and Qumulo; Cohesity, Commvault, HPE, NetApp, Nutanix, Rubrik, and VAST Data are named for future availability.
Industry backing
Support at announcement is broad. OpenAI Head of Enterprise Product Alexander Embiricos is quoted: "We believe in open AI ecosystems and are excited to collaborate with Databricks on providing a standard, secure way to discover and authorize access to AI assets." Enterprise adopters named at launch include SAP (Business Data Cloud), Stripe (Data Pipeline), Atlassian (Analytics), LSEG, and Amadeus. Linux Foundation CEO Jim Zemlin stated: "OpenSharing addresses a critical need for a common, vendor-neutral framework that enables organizations to exchange AI assets securely and interoperably across platforms and ecosystems." Databricks co-founder and CTO Matei Zaharia is quoted: "Delta Sharing proved the industry would choose open over locked-in. OpenSharing extends that principle to the full AI stack, while expanding the cross-platform ecosystem to Iceberg recipients and on-premises providers."
Why it matters
Industry pattern: open sharing protocols have historically expanded ecosystem reach while reducing vendor lock-in - Delta Sharing's adoption across Databricks, Snowflake, Oracle, and major BI tools demonstrated this dynamic. OpenSharing attempts to establish the same open-standard foundation for AI asset exchange before proprietary agent marketplaces become entrenched. For data engineers and ML practitioners building agentic systems, a standard protocol for distributing agent skills and models across organizations removes the need for custom point-to-point integrations with each partner. The project is hosted at the Linux Foundation - the same governance model used by the Linux kernel, Kubernetes, and other widely adopted open infrastructure projects - which lends vendor-neutral credibility to the effort.
Scoring Rationale #
OpenSharing is a Linux Foundation-hosted open standard with broad industry backing (OpenAI, SAP, Stripe, Atlassian, LSEG, Amadeus) that addresses a genuine infrastructure gap for the agentic era: no prior open protocol existed for sharing agent skills and AI models across organizations. The combination of vendor-neutral governance, Iceberg support, on-premises reach, and major partner adoption at launch makes this a notable data infrastructure event for ML and data engineering practitioners.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.