AI Spending Rises as 53% of Organisations Struggle to Embed Business Context
Organisations are increasing their investment in artificial intelligence, but many are struggling to provide AI systems with the business context needed to turn that spending into measurable outcomes, according to new research from Alteryx.
The company’s 2026 IT Leader Research: The State of AI Ownership, Agents, and ROI surveyed 1,400 IT leaders globally and found that 80% expect AI spending to increase over the next two years, while 69% already report moderate or significant returns from their AI investments.
However, 53% said their organisations struggle to translate business context into AI systems and workflows, exposing a potential barrier to scaling AI beyond individual use cases.
The findings also highlight the importance of business knowledge in AI deployment. While 77% of technology leaders agree that business context is critical to producing accurate and relevant AI outputs, much of the knowledge required by AI systems remains embedded in spreadsheets, documentation, email exchanges and the experience of employees.
AI investment moves towards measurable outcomes
As organisations shift from AI experimentation towards broader deployment, expectations around returns are also increasing. Technology leaders are primarily measuring AI success through productivity improvements, cited by 53% of respondents, followed by cost reduction at 45% and revenue growth or wider business impact at 39%.
Meanwhile, 35% said the ability to measure AI ROI will be among the capabilities that most distinguishes technology leaders from their peers.
The research suggests that the focus of enterprise AI is moving from whether the technology can work to whether organisations can consistently convert it into measurable business value.
Business logic remains difficult to operationalise
Alteryx said organisations face a challenge beyond simply providing AI with more data. AI systems also need access to the rules, definitions, policies, thresholds and decision criteria that determine how individual businesses operate.
“Our research highlights a growing gap between AI ambition and enterprise-scale execution,” said Andy MacMillan, CEO of Alteryx. “Organizations have proven they’re willing to invest in AI, and many are already seeing returns. But scaling AI requires more than better models. It requires making the business knowledge people use every day available to the systems making decisions.”
The issue becomes particularly important as organisations deploy agentic AI systems that can increasingly take action rather than simply generate information. Some 93% of surveyed IT leaders said they are confident agentic AI could deliver measurable ROI for their enterprise within the next two years.
Limited access to data remains a barrier
The research also found that only 18% of organisations have achieved fully self-service access to cloud data for business users. Thirty-eight percent operate a mixed model, while 15% said business users remain largely dependent on technical teams for data access and analytics.
This dependency can create a disconnect between the employees who understand business processes most deeply and the teams responsible for developing AI systems, making it harder to incorporate operational knowledge into AI workflows.
IT-business collaboration becomes more important
The findings point to increasing recognition that successful AI deployment requires closer collaboration between technology and business teams. Two-thirds of technology leaders said AI and agent-based systems are most productive when managed within the line of business, while 71% believe AI initiatives are most successful when IT and business teams collaborate closely.
Despite this, strategy and delivery remain concentrated within IT, at 37% and 38% respectively, while business teams are most commonly responsible for defining requirements, at 30%.
MacMillan concluded, “The organizations creating lasting value from AI will be the ones that operationalize their business logic so it becomes visible, governed, repeatable, and ready for AI.”
The research was conducted by Coleman Parkes in April and May 2026 among 1,400 technology leaders and automation specialists across North America, Europe, the Middle East and APAC. Respondents represented sectors including banking, manufacturing, retail and consumer goods, insurance, and public sector and education organisations.
