IBM Closes $11 Billion Confluent Acquisition, Absorbing Apache Kafka's Commercial Steward Into Its AI Data Stack
IBM completed its largest acquisition since Red Hat, buying data streaming company Confluent for $31 per share and delisting it from Nasdaq as it bets on real-time data as the backbone of enterprise AI.
Overview
IBM completed its $11 billion all-cash acquisition of Confluent on March 17, 2026, absorbing the company that commercialized Apache Kafka into a software division now squarely focused on feeding real-time data to AI agents and automated workflows. At $31 per share, the deal represents a 34 percent premium over Confluent’s pre-announcement stock price and an enterprise value of roughly $11 billion, according to PR Newswire. Confluent’s Class A common stock has been delisted from the Nasdaq, ending the company’s run as a public entity.
The acquisition is the largest IBM has made since its $34 billion purchase of Red Hat in 2019 and follows its $6.4 billion HashiCorp deal in 2024, as reported by The Register. Together, the three deals form an infrastructure stack spanning containers, infrastructure as code, and now data streaming, all unified under IBM’s hybrid cloud and AI strategy.
What We Know
Confluent serves more than 6,500 enterprises, including over 40 percent of the Fortune 500, with a platform built on Apache Kafka that enables real-time data streaming across cloud and on-premises environments, according to PR Newswire. The company was co-founded by Jay Kreps, who created Apache Kafka at LinkedIn before spinning it out as a commercial venture, as noted by TechCrunch.
IBM CEO Arvind Krishna framed the deal as enabling a “smart data platform for enterprise IT, purpose-built for AI,” according to The Register. The strategic thesis centers on a gap IBM sees in production AI deployments: most enterprise data remains siloed and hours or days old by the time it reaches AI models, making real-time streaming infrastructure essential for the agentic AI systems IBM is betting on.
Three day-one integrations went live immediately upon closing, according to PR Newswire:
- watsonx.data now streams live operational events so AI models run on continuously updated data with lineage and policy enforcement.
- IBM Z mainframes can identify and stream real-time transactional data directly into analytics and AI workflows.
- IBM MQ and webMethods extend event-driven automation with high-scale streaming across hybrid environments.
Confluent’s stockholders approved the merger on February 12, 2026, with shareholders controlling approximately 62 percent of voting power having pledged support prior to the meeting, as reported by The Register. IBM projects the acquisition will boost adjusted EBITDA within its first year and enhance free cash flow by year two.
What We Don’t Know
The deal raises several unanswered questions that will shape its long-term significance.
The most consequential concerns the future of Apache Kafka’s open-source ecosystem. While IBM has publicly stated that its acquisition rationale depends on Kafka’s openness and broad adoption, the company’s track record with open-source acquisitions is mixed. IBM’s earlier absorption of HashiCorp has already produced friction, with the March 31 elimination of HCP Terraform’s legacy free tier accelerating migration toward OpenTofu and other alternatives.
Reports have surfaced of significant layoffs among former Confluent employees following the close, though neither IBM nor Confluent has confirmed specific numbers. The impact on Kafka Improvement Proposals (KIPs) and the broader contributor community remains unclear. Other companies with commercial interests in the Kafka protocol, including AWS, Aiven, and Redpanda, maintain their own investments in the project, which may buffer against any reduction in Confluent’s open-source contributions.
Confluent’s addressable market had doubled to $100 billion since 2021, according to The Register, yet the company had not yet reached profitability. Whether IBM can convert that market opportunity into profitable growth while maintaining Confluent’s developer relationships and community trust will determine the deal’s success.
Analysis
The Confluent acquisition marks a shift in IBM’s strategy from assembling infrastructure components to building a vertically integrated AI data supply chain. Where Red Hat gave IBM the container and operating system layer and HashiCorp provided infrastructure automation, Confluent fills the real-time data plumbing that connects them. The pattern suggests IBM views itself not as a cloud hyperscaler competing directly with AWS, Azure, or Google Cloud, but as the middleware and data governance layer that enterprises need to operate across all of them.
The timing reflects a broader industry consensus that batch-processed, warehouse-centric data architectures are insufficient for production AI systems that need to act on current operational signals. Confluent’s proprietary Kora engine, which provides cloud-native multi-tenant streaming beyond what open-source Kafka offers out of the box, gives IBM differentiated technology that hyperscaler alternatives like AWS Kinesis, Azure Event Hubs, and Google Pub/Sub have not fully replicated.
For Kafka users outside the IBM ecosystem, the near-term risk is concentrated in Confluent’s commercial and proprietary layers rather than the Apache project itself. Managed connectors, enterprise licensing terms, and Confluent Cloud pricing are the areas most likely to see changes under IBM ownership. Organizations with deep dependencies on Confluent-specific tooling may want to evaluate their lock-in exposure, while those running self-managed Kafka clusters will likely see little immediate impact.