AI doesn't fail organisations – fragile culture, skills and data foundations block real impact
At the start of 2025, 32% of HR leaders predicted AI would be transformative for their organisations. Twelve months later, only 15% said it delivered on that promise, according to ELMO's 2026 HR Industry Benchmark Report.
Before anyone blames the technology, it’s worth asking what was actually in place before the rollout began.
According to ELMO’s research, 35% of HR leaders say a lack of skilled people is the biggest barrier to AI adoption, with a lack of expert guidance close behind at 33%. Budget and time constraints rank lower. The skills gap is significant and it’s not something technology alone can fix.
The foundations problem
Many HR teams already have people who are genuinely capable with AI. The challenge is the environment they’re working in: data fragmented across multiple systems, metrics requiring manual extraction, insights that are hard to act on. When those foundations are missing, AI gets used for small efficiency wins rather than tasks such as assisting with strategic decision-making.
The organisations seeing the strongest results are the ones that prioritised readiness before scale. They connected core data sources. Clarified which workforce metrics mattered most to leadership. Made sure HR could access those insights quickly. That kind of connectivity rarely happens when HR data is scattered across disconnected systems. This is still the reality for most, with only 23% of organisations having a fully centralised platform across HR, payroll and workforce management (HRIB 2026). Once that groundwork is in place, AI shifts from being a summarisation tool to helping leaders model workforce scenarios, forecast capability gaps, and make faster decisions. Not just faster admin.
Foundations first. Then scale.
The ownership problem
Senior leaders are most likely to say AI is owned by the C-Suite (39%) or IT (35%). That ambiguity is expensive. Everyone expects progress. No one is empowered to drive it.
A clear responsibility map fixes this. IT owns the infrastructure - data, systems, security. HR leads the human side - capability building, governance, behaviour change, impact measurement - with explicit C-Suite sponsorship. Only 1 in 4 Australian HR leaders feel fully equipped to meet leadership’s expectations around AI, yet leadership rates HR more positively than HR rates itself. That gap is a lever. Use it to make the case for resources by building on demonstrated strengths, not cataloguing what’s missing.
The anxiety nobody is addressing
Two in five employees are now concerned their role could be made redundant, up from 36% the previous quarter, according to ELMO’s latest Employee Sentiment Index. Half say they feel pressure to work harder or longer just to feel secure in their job. An AI strategy that ignores those fears will struggle to gain real traction - people experiment when they feel safe, and shut down when they don't.
Building AI-fluent, un-anxious employees requires more than a training session or two. It needs a cultural environment where experimentation is genuinely encouraged, skill-building is ongoing, and leaders are visibly doing the same learning they’re asking of everyone else.
The leaders in AI aren’t defined by the tools they’ve bought. They got their data, systems, and people in order first - and created the conditions where their people actually wanted to use them.
ELMO is the Complete AI Workforce Platform™. Find out more here.
This article was produced in partnership with ELMO Software
