Real-Time Data Infrastructure Hinders AI Adoption

Nearly three-quarters, or 72%, of global IT leaders say a lack of real-time data infrastructure is the main obstacle to expanding their artificial intelligence (AI) initiatives. The finding comes from the 2026 Data Streaming Report by Confluent, which surveyed 4,625 IT leaders worldwide, including respondents from Indonesia. The study examined the challenges companies face when developing AI and why organisations need to focus more on improving foundational infrastructure rather than simply increasing investment in AI models themselves.

According to the research, 72% of IT leaders face at least three significant challenges when expanding AI initiatives. These infrastructure issues also directly impact the development of autonomous AI. A total of 66% of IT leaders cited data quality and infrastructure as the primary barriers. As a result, only 32% of respondents reported having deployed autonomous AI in a production environment, with the majority still experiencing delays.

As AI transitions from pilot projects to production, industry attention is shifting to the data that powers it. A total of 80% of IT leaders stated that using company data to drive AI-based systems is a top business priority. The findings indicate that organisations view data streaming as a crucial solution. Approximately 88% of respondents said that data streaming platforms help pave the way for autonomous AI by making data more trustworthy, contextual, and discoverable. Furthermore, 94% believe data streaming will enhance the impact of their AI investments.

Interestingly, investment in data streaming is now beginning to surpass investment in AI itself. A total of 88% of IT leaders rank data streaming as a top priority, compared to 82% for AI and machine learning technologies. This signals a growing awareness that the value of AI is highly dependent on access to reliable, real-time data.

Shaun Clowes, Chief Product Officer at Confluent, explained that many organisations do not have an AI investment problem but rather a data problem. “AI systems depend on current and accurate information. Yet too many systems are still built on fragmented data and batch processes that were not designed for continuous intelligence,” he said.

This pattern of infrastructure barriers is even more pronounced in Indonesia. A total of 82% of Indonesian IT leaders reported facing at least three major challenges in AI adoption. Of these, 79% specifically highlighted inadequate real-time data infrastructure as a barrier to progress. Despite this, Indonesian IT leaders showed the strongest confidence globally that data streaming can close the gap. Nearly all respondents (99%) expect data streaming to increase the impact of their AI investments, and 98% believe the technology accelerates AI adoption within their companies.

Rully Moulany, Area Vice President Asia at Confluent, explained that organisations in Indonesia have big ambitions but are held back by legacy infrastructure. “IT leaders in Indonesia are more convinced that data streaming can close this gap. That is why the technology has overtaken AI as the top strategic priority in Indonesia,” he concluded.