IBM has entered into a definitive agreement to acquire data streaming leader Confluent, a move designed to cement its infrastructure capabilities for the generative AI era. In a memo sent to employees on December 8, 2025, IBM Chairman and CEO Arvind Krishna described the acquisition as a “decisive step” that accelerates the company’s hybrid cloud and AI strategy by establishing an “intelligent, always-on core” for enterprise data.
The acquisition is predicated on the industry’s shift from static data storage to real-time data movement. Krishna emphasized that as companies increasingly deploy AI agents, digital workers, and autonomous workflows, the need for “live” data signals has become critical. “Models are only as strong as the signals feeding them,” Krishna wrote, arguing that modern operations—ranging from millisecond transactions to instant supply chain adjustments—require an infrastructure capable of moving data immediately from its source to where it is needed.
By integrating Confluent, which specializes in real-time data streaming, IBM aims to complete the foundation of what it calls an “end-to-end Smart Data Platform”. This platform is intended to unify batch processing systems with real-time streams, allowing clients to “move, manage, and act on data” across complex hybrid environments. Krishna highlighted that this capability opens one of the “largest new infrastructure opportunities of the AI era,” as autonomous systems require continuous access to data to function effectively.
The deal also reinforces IBM’s strategy of absorbing category-defining open-source technologies to strengthen its enterprise stack. Krishna drew direct parallels to the company’s previous acquisitions of Red Hat and HashiCorp, noting that Confluent fits the same “disciplined playbook” of identifying and scaling powerful open-source platforms globally. While financial terms were not disclosed in the internal announcement, the move signals a continued reliance on targeted M&A to bolster IBM’s core pillars of automation, security, and integration.
