More than half of enterprise AI projects are failing to deliver their objectives, with poor unstructured data management identified as the primary blocker, new Nasuni research has found.

Some 97 per cent of organisations have deployed or are piloting AI agents, according to The State of Enterprise File Data Annual Report 2026. However, 57 per cent of AI projects are not delivering on their stated objectives.

Nearly all enterprises (94 per cent) are struggling to manage unstructured data, which comprises the majority of their data footprint, the report found.

Only 16 per cent currently prioritise unstructured data management as a core IT investment. However, 60 per cent plan to invest in this area over the next 18 months.

“Enterprises are moving fast on AI projects, but most aren’t getting the results they want,” said Sam King, chief executive officer at Nasuni.

“What this report makes clear is that AI success depends on how well you manage and prepare your data,” King said.

“Too many organizations are still relying on outdated approaches to unstructured data management, limiting their ability to unlock its full value,” he said.

Some 90 per cent of organisations reported barriers to scaling AI. Top barriers included data security concerns (43 per cent), integration roadblocks (36 per cent) and lack of trust in data (33 per cent).

Nearly half (46 per cent) of organisations said AI initiatives had revealed issues with data quality and governance. A further 79 per cent reported inconsistent file access and performance across locations.

While nearly all organisations are piloting AI agents, only 18 per cent have deployed them at scale. The findings point to a disconnect between AI ambition and operational readiness.

Some 62 per cent of organisations expect hardware costs to rise as key components such as DRAM surge in price. Rising costs are compounding pressure on the infrastructure needed to support data-intensive AI workloads.

Industry-specific concerns vary across sectors. In architecture, engineering and construction, 66 per cent of firms cited security as their top unstructured data concern.

Manufacturers continue to face elevated cyber risk and longer recovery times. Energy, oil and gas organisations remain split on whether AI decision-making should sit with the C-suite or IT.

The Sapio Research survey polled 1,000 purchasing decision-makers in March 2026 at organisations with 1,000 or more employees. It covered the US, UK, France and the DACH region of Germany, Austria and Switzerland.

The full report is available here.