Where AI is and isn’t helping workplaces

Skills shortages, shifts in hiring patterns, and confusion around AI guidelines and ownership are plaguing Australian organisations.

New research from ELMO Software reveals that AI is not delivering at the rate it is being adopted, thanks to higher rates of employee retention and skills shortages not being addressed.

According to the 2026 ELMO HR Industry Benchmark Report, AI continues to sit in a guideline grey zone, and as a result, there is a clear disconnect in Australian workplaces. While 32 per cent of HR leaders anticipated organisational transformation from AI, only 15 per cent said it actually delivered.

Based on a survey of 904 HR professionals in January 2026, the findings revealed how AI intent continues to be higher than actuality.

While AI adoption increases, the growth is anything but uniform. Currently, four in five Australian organisations and 93 per cent of HR teams reportedly use AI; however, more than half of the HR professionals use it only occasionally.

In addition, just 6 per cent of surveyed organisations report that more than three-quarters of the workforce use approved AI tools.

According to ELMO Software president Joseph Lyons, the findings reinforce the tension between adoption and impact.

“Occasionally using an AI tool or a large language model will not lead to true organisational transformation. We’re seeing a capability gap between AI readiness and actual impact. The focus now is moving beyond standalone tools and embedding AI into the systems that support how people are hired, onboarded, developed and managed,” Lyons said.

Lyons also highlighted that organisations should take advantage of the increasingly rampant “job-hugging” trend, in which organisations are retaining employees for longer as workers prioritise financial security.

He said: “As employees stay longer in their roles, organisations have a greater opportunity to invest in upskilling and career development. Embedding AI capability and continuous learning into onboarding will be critical to helping employees adapt as roles and workflows evolve.

“That’s exactly why organisations are starting to think about AI not as a standalone tool, but as part of a connected workforce platform that supports the full employee life cycle.”

Organisations that capitalise on employee retention trends may be poised to fare the uncertain weather better. Turnover rates have dropped for the first time in four years, hitting a low of 11 per cent. At the same time, the average cost to hire is now $17,000, down 9 per cent compared to last year.

More importantly, it’s taking an extra five days compared to last year (25 from 20) to fill vacant roles and a further 44 days for new hires to become productive, compared to 34 days in 2025.

Lyons said: “Many organisations are focusing on deploying AI tools, but the real challenge is workforce capability.”

The issues don’t stop here: confusion around AI ownership is also causing issues. Only 12 per cent of HR professionals believe they are in charge of managing AI adoption, while 39 per cent believe the power sits with IT. In addition, 39 per cent of business leaders believe the responsibility lies with the C-Suite or IT, but still look to HR for results.

Just over a quarter of HR feel fully prepared to deliver on these AI expectations.

Dr Amantha Imber, ELMO partner, organisational psychologist, and AI Consultant of the Year, explained that this disconnect again stalls progress. “Everyone expects progress, but no one is truly empowered to drive it end to end. The fastest path forward is a clear responsibility map,” she said.

According to the report, successful translation of AI usage to outcomes is held back by two main barriers within compliance and infrastructure: 37 per cent referencing data and privacy concerns, and 34 per cent fearing potential issues integrating AI into current systems.

Trust remains to be a key issue, with around a third of HR professionals citing challenges with excessive rework, and more highlighting difficulty validating outputs, inaccuracy, and AI actually slowing tasks down.

Lyons said: “If businesses don’t fix foundational issues around data quality and integration, AI will continue to underdeliver.”

“Addressing these first will go a long way to properly equip organisations to turn adoption into strategic impact.”