OpenSearch Gains Ground as Organizations Build AI Infrastructure Beyond Search

The trajectory of OpenSearch reflects a striking shift in how the project is being used. What started in 2021 as AWS's response to Elastic's licensing change—a fork maintained under the Apache 2.0 license—has evolved into a broader platform for data infrastructure. The Linux Foundation established the OpenSearch Software Foundation in September 2024 to provide vendor-neutral governance, marking a turning point in the project's maturity.

To understand how organizations are deploying OpenSearch today, the Linux Foundation Research team and the OpenSearch Software Foundation jointly released The 2026 Open Data Infrastructure Report. The survey gathered responses from 294 participants representing IT vendors, end-user organizations, and independent consultants during May 2026.

Key findings from the report

Generative AI and large language model applications dominate the AI landscape, cited by 82% of surveyed organizations. Across all regions, at least 83% of respondents are either running AI workloads or planning to do so. Among those already using OpenSearch, 62% have incorporated it into AI workloads, and 44% view it as a core component of their AI infrastructure rather than a secondary tool.

Awareness of OpenSearch has grown substantially. In 2026, 89% of organizations had heard of the project, compared to 68% two years earlier. Production deployments increased from 19% to 36% during the same timeframe, with an additional 36% currently testing or evaluating the platform.

When selecting a data infrastructure platform, organizations prioritize several factors. Total cost of ownership ranks as the leading criterion for 80% of respondents, followed by security and compliance at 79%, and vendor independence at 69%. Notably, 71% consider running infrastructure outside a single cloud provider's ecosystem a strategic imperative, and the median annual data infrastructure budget across the sample reaches approximately $2.4 million.

Current OpenSearch deployments serve multiple purposes. Search and retrieval operations lead at 91% of users, trailed by log analytics and observability at 83%, real-time analytics at 70%, and AI-related workloads at 62%. Hybrid search approaches have become the standard, adopted by 68% of organizations.

Adoption remains early stage

Despite growing interest, most OpenSearch implementations have not yet reached critical importance within their organizations. Sixty percent of active users characterize their deployments as experimental or replaceable, while only 36% describe them as deeply embedded or mission-critical infrastructure.

Among organizations not currently using OpenSearch, 81% express openness to evaluation, though only 14% have concrete plans to pursue adoption.

The foundation has announced three new members joining the OpenSearch Software Foundation, including Intel and two additional long-term support vendors.

Bianca Lewis, Executive Director of the OpenSearch Foundation, offered perspective on how the platform fits into the emerging landscape of agentic AI:

The technology stack powering agentic AI is shifting rapidly, and with growing uncertainty around AI costs, long-term commitments to proprietary platforms present both a financial and architectural risk. Now, leaders are turning to an open, vendor-neutral platform to strategically mitigate risk.

Bianca Lewis, Executive Director of the OpenSearch Foundation

An open data layer provides cost data and full visibility into which services are being called, giving teams the clarity needed to accurately cost and manage their infrastructure. This delivers the flexibility to innovate without vendor lock-in, ensuring organizations maintain sovereignty over both their data and their spend as the future of AI takes shape.

Bianca Lewis, Executive Director of the OpenSearch Foundation