Australian AI readiness drops: Survey

Organisations are struggling to scale AI adoption as risks and infrastructure pose challenges.

Image:
Graeme Beardsell, MD ANZ, Kyndryl

AI is putting pressure on business models, technology foundations and operations, according to the latest Kyndryl Readiness Report, which surveyed 3,700 senior business leaders across 21 countries, including Australia.

Organisations are looking for returns on their significant AI investments, yet they’re finding it more difficult to manage the implications of incorporating AI into business operations. Over a quarter (28 percent) saw an AI investment fail, while 25 percent cancelled a major AI initiative.

“As AI becomes more deeply embedded in critical business operations, governance, security and digital sovereignty are becoming increasingly important considerations,” said Graeme Beardsell, MD ANZ at Kyndryl.

Around two-thirds (64 percent) of respondents say their infrastructure requires significant re-architecture before it can support AI at scale while average readiness has fallen 10 points to 23 percent since last year’s report.

The steepest declines were in readiness for regulatory change (down 18 points to 18 percent) and competitive disruption (down 16 points to 22 percent).

“Many Australian organisations underestimate the level of preparation needed to manage the risks associated with AI adoption,” said Beardsell.

Locally, nearly nine in ten (89 percent) are deploying AI in workflows beyond pilot projects, compared with 84 percent globally. And fewer say innovation efforts are delayed by foundational issues in their technology stack than a year ago (down 12 points to 51 percent).

The report showed some encouraging results, with 65 percent of Australian leaders reporting positive returns on AI/ML investments, up 8 points from 2025. However, 66 percent also say they have more technology pilot projects than they can realistically scale.

Beardsell said these figures suggest the next phase of AI adoption will require greater discipline around investment decisions.

“Rather than pursuing every new opportunity, organisations need to identify where the technology can deliver meaningful business outcomes and ensure they have the capabilities to support those investments over time,” he said.

“That includes managing AI consumption and associated costs, including token usage, to avoid unexpected expenses as deployments scale,” he added.

Security is feeding into lower AI readiness

For the third year in a row, security concerns topped the list of external risks globally, and this year contributed most to the overall decline in global readiness.

In Australia, cyber readiness dropped 13 points to 28 percent in 2026 and 61 percent of respondents say cybersecurity risks are growing faster than their ability to address them, compared with 52 percent globally.

Beardsell said organisations need to manage emerging risks, protect sensitive data and ensure the complex systems underpinning their businesses continue to operate securely and reliably.

“Scaling AI is not just a technology challenge. Organisations also need to consider how AI changes the way people work, ensuring employees have the right skills, training and support to adapt as new technologies become embedded in everyday operations,” he told CRN Australia.

“The real measure of success won’t simply be how quickly organisations adopt AI, but whether they can turn early returns into repeatable business value without compromising security, resilience or operational control,” he added.

Highlights