In artificial intelligence, around 72% of IT leaders globally report challenges with real-time data infrastructure, according to the 2026 Data Streaming Report released by Confluent. The report is based on input from 4,625 IT professionals worldwide and examines infrastructure-related issues that can affect the scaling of AI initiatives. It also highlights a focus among respondents on strengthening underlying data systems rather than solely increasing investment in AI.
A reported 72% of IT leaders identify limitations in real-time data processing as a key barrier. Additional challenges include uncertainty around data lineage, timeliness, and quality, cited by 66%, as well as fragmented data ownership, noted by 65%.
The report also links these infrastructure factors to delays in deploying agentic AI systems. Approximately 66% of IT leaders attribute slower adoption of agentic AI to data quality and infrastructure issues, while 32% report having successfully integrated agentic AI into production.
As organisations move AI projects from experimentation to production use, attention is increasingly placed on the quality and timeliness of data used in these systems. Around 80% of IT leaders identify the use of enterprise data in AI systems as a top business priority, reflecting the importance placed on real-time data access.
Data streaming platforms are widely viewed as a significant component of data infrastructure. Nearly 88% of leaders report that these platforms improve data trustworthiness and accessibility. In addition, 94% expect data streaming to have a greater impact on AI investments, while 90% consider it important for enabling AI adoption.
The report notes that investment in data streaming is increasingly comparable to investment in AI and machine learning technologies, cited by 88% and 82% of IT leaders respectively. As organisations shift AI initiatives toward operational use, attention to data infrastructure is becoming more prominent. The findings suggest that the effectiveness of AI systems is closely linked to the availability of reliable, real-time data, leading many IT leaders to prioritise investment in data foundations.
Overall, the results indicate that while AI models remain central to development efforts, the supporting data infrastructure plays a key role in determining how effectively AI investments translate into operational outcomes.
