66% of individuals may be using AI, but 79% are wary of the risks, finds KPMG

Two-thirds of individuals are using AI regularly, with 33% using it weekly.

That’s according to new research from KPMG and the University of Melbourne.

The pair surveyed 48,000 individuals in 47 countries, and found that 83% believe that AI usage will result in a wide range of benefits, including greater efficiency (67%), more access to information (67%) and higher quality work (58%).

Plus, 46% said they were using AI to increase revenue generating activities.

Despite this widespread AI adoption, KPMG and the University of Melbourne found that trust remains a critical challenge with just 46% of respondents saying they trusted AI tools.

Speaking exclusively to UNLEASH about the report’s findings, Nhlamu Dlomu, Global Head of People, KPMG International, shares: “One of the most interesting trends in the study is the dichotomy between people’s excitement and desire to use AI on one hand, but also their lack of trust and wariness of the risks on the other.

“For example, we saw contrasts in trust and concern, with 68% of respondents feeling optimistic about AI, while 54% were wary of trusting it.

“Similarly, 73% of respondents reported experiencing the benefits of AI, but 79% were also concerned about possible risks.

“With all changes as transformational as AI, I think this mix of excitement and anxiousness is natural.”

In terms of the negatives of AI, the respondents shared that AI actually increased their workload and added to their stress (26%), led to repetitive work (39%), and caused issues with compliance (35%).

How HR can build trust around AI

UNLEASH was keen to dig into the data more to figure out the solution to the ongoing trust issue with AI.

In fact, KPMG and University of Melbourne compared trust data in 2025 to 2022 and found that it had declined from 63% to 56%, while the proportion of people worried AI increased from 49% to 62%.

According to KPMG’s Global Head of AI, David Rowlands, “organizations have a clear role to play when it comes to ensuring that AI is both trustworthy and trusted”.

They need to create “a culture of responsible, open and accountable AI use”, Professor Nicole Gillespie from the University of Melbourne noted.

Part of the solution is creating guidelines and governance around AI use at work.

This is urgent need as the KPMG-University of Melbourne data found that 47% of individuals admit using AI in ways that could be deemed inappropriate and two-thirds admit relying on AI output without evaluating its accuracy.

Another element is a focus on training and educating employees around AI – just 47% report training in AI or related fields.

KPMG’s Dlomu notes: “All employees, from the most senior to the most junior, will need to learn about AI and how to harness its potential.

“HR leaders need to ensure their people have access to relevant learning and development opportunities alongside building a culture that supports and encourages continuous learning.

“We don’t yet know the full potential of AI, so organizations that also put an emphasis on creating safe spaces for questions and experimentation will be at an advantage.”

As a result, Dlomu adds that HR needs to step up to support leaders.

We’re putting a lot of demands on leaders – HR leaders need to ensure they have the support they need.”

AI is top of the agenda at UNLEASH America next week – it’s not too late to grab a pass and join us in Las Vegas.