Only 8% of firms track AI impact, Talent Group says

Talent Group has published research showing that only 8% of organisations in Australia and New Zealand can track the impact of artificial intelligence using clear metrics or KPIs. The findings are based on a survey of 1,505 business leaders and technology professionals across more than 12 industries.

The report points to a sharp rise in AI adoption over the past year. It found that 53% of organisations now use AI in workflows or embed it in strategy, up from 13% a year earlier. The share of businesses that have fully embedded AI into both strategy and operating models rose from 4% to 24%, suggesting implementation is moving beyond trial phases even as measurement remains limited.

A central theme in the data is the gap between deployment and proof of value. While 85% of individual respondents said AI had made them more efficient at work, only a small minority of organisations said they were measuring outcomes through formal performance indicators.

That contrast highlights a familiar challenge for executives under pressure to show returns on technology spending. Employees may report completing tasks faster or carrying lighter workloads, but without agreed measures linked to cost, quality, risk or customer outcomes, businesses may struggle to show whether those gains translate into broader operational or financial results.

Adoption rises

The figures suggest AI has become a mainstream workplace tool in parts of the Australia and New Zealand market. The jump from 13% to 53% in one year among organisations using AI in workflows or strategy indicates a notable shift in how quickly companies are integrating the technology into day-to-day operations and planning.

At the same time, the rise from 4% to 24% in organisations that say AI is fully embedded in strategy and operating models points to a more structural change. Rather than confining AI to isolated pilots, some employers appear to be building it into how teams work, make decisions and organise processes.

Yet the report suggests adoption has outpaced the systems needed to assess whether those decisions are paying off. The low level of KPI-based tracking means many organisations may be relying on anecdotal evidence or usage figures rather than business results.

Jack Jorgensen, General Manager - Data, AI & Innovation at Avec, said that approach can be misleading.

"The more precisely you measure AI usage, the more precisely you risk measuring the wrong thing," Jorgensen said.

"Stop asking how much AI people are using and start asking whether the outcomes the business already cared about - speed, quality, cost, risk and customer experience - are actually moving. If ROI becomes an adoption target, teams optimise for the metric, not the return."

Governance concerns

The research also identified weaknesses in governance and staff awareness. Only 16% of respondents said they were confident staff know what data is safe to input into AI tools, raising questions about how employers are managing privacy, confidentiality and internal controls as use broadens.

According to the survey, nearly one in four organisations have never provided AI training or refreshed internal policy. That suggests some businesses are allowing staff to adopt the technology faster than managers are updating rules or guidance.

Those findings are likely to resonate with companies facing increasing scrutiny over how employees use consumer and workplace AI systems. Risks can range from the disclosure of sensitive information to poor-quality outputs or decisions made without sufficient human review.

Jorgensen said the data pointed to a practical rather than theoretical risk.

"These figures are deeply concerning. If only 16% are very confident that people understand what data is safe to enter into AI tools, and 1 in 4 have never received AI training or a policy refresh, the risk isn't theoretical. Without stronger governance and practical training, organisations will likely see more incidents involving sensitive data, misuse of tools and avoidable exposure."

Management challenge

For business leaders, the report presents a mixed picture. On one hand, the rapid spread of AI in workflows suggests organisations see enough immediate value to keep expanding its use. On the other, the lack of measurement and weak governance structures indicate many boards and management teams still do not have a clear framework for judging success or controlling risk.

The survey spans business leaders and technology professionals, giving it a view across both strategic and operational decision-makers. That makes the contrast between perceived efficiency gains and limited formal measurement more notable, because it suggests the issue is not simply technical implementation but management discipline.

Talent Group describes itself as a multi-brand business spanning recruitment, workforce solutions, consulting, IT project delivery, workforce augmentation and social impact, with operations across Australia, New Zealand and the US. Its specialist brands include Talent, Solve by Talent, Avec, Scale by Avec and Rise by Talent.

The clearest finding in the research remains the gap between the 85% of individuals who say AI has improved their efficiency and the 8% of organisations that can prove its impact with clear metrics or KPIs.