Australian firms see AI IT gains but still demand oversight

Riverbed has published a global survey showing that Australian organisations are seeing returns from using AI in IT operations. The findings also show that most still want human approval before AI makes operational decisions.

The study surveyed 1,200 business decision-makers, IT leaders and technical specialists across seven countries. In Australia, 93% of respondents said AI in IT operations was meeting or exceeding return on investment goals, ahead of the 91% global figure.

That positive assessment sits alongside a more cautious view on autonomy. In Australia, 84% of respondents said their organisation was hesitant to let AI make operational decisions without human approval, compared with 77% globally. Only 14% expected full autonomy within a year, versus 21% worldwide.

Globally, 90% of organisations want to use agentic AI to support autonomous IT operations. Meanwhile, 76% view agentic AI as central to future IT strategy, while 92% believe AI observability and governance will become a critical IT domain.

Readiness gap

The results point to a gap between executive ambition and operational readiness. Only 41% of organisations globally consider their environments fully prepared for AI, while on average just 19% of IT operations are automated today.

Adoption at scale also remains limited. Only 8% reported enterprise-wide AI-enabled operations, and 9% said AI initiatives were fully deployed across their organisations.

Australian respondents highlighted workforce and systems issues as barriers to wider use. Skills gaps were cited by 45%, above the 37% global figure, while 95% said consolidation would reduce friction in AI adoption.

These findings align with the survey's broader emphasis on visibility, data quality and simpler infrastructure. Globally, 96% said multi-domain visibility was important for AI-driven IT operations, but only 17% said they had fully unified visibility across networks, applications, endpoints and cloud environments.

Riverbed also found weak confidence in the underlying quality of information used by AI systems. Just 23% rated their data granularity as excellent, and only 21% said the same of data quality.

Governance focus

The survey suggests governance is becoming a central concern as companies move AI from trials into day-to-day IT management. Only 42% of respondents said they were highly confident in their organisation's AI governance, with that share falling to 36% among technical specialists.

Trust in AI recommendations also remains limited. Just 24% said they extensively trust AI recommendations with limited human review.

Security and operational risk were among the most commonly cited obstacles. Some 55% of organisations named security and compliance concerns as a top challenge, followed by the risk of operational disruption at 45% and lack of trust in AI decisions at 39%.

The research also examined how AI may reshape service desk work and digital employee experience. Nearly all respondents, 93%, said AI-powered operational intelligence would transform frontline IT and service desk functions within two years.

By 2028, 27% predicted autonomous operations would significantly reduce service desk activity, 22% said agentic AI would allow teams to focus on more complex issues, and 35% expected service desk roles to shift from reactive support to proactive prevention.

Even so, present-day use appears limited. Only 34% said their IT operations teams were extremely effective at using AI to identify and resolve issues before employees noticed a disruption.

Jim Gargan, Chief Marketing Officer at Riverbed, said the research showed a divide between interest in autonomous IT and the conditions needed to support it.

"The journey to autonomous IT is well underway. Organisations increasingly see AI as the path to more resilient, efficient and proactive IT, but our research shows significant gaps between ambition and readiness.

"While 96% say multi-domain visibility is important for AI-driven IT operations, only 17% have fully unified visibility across networks, applications, endpoints and cloud. At the same time, 92% believe AI observability and governance will become a critical new IT domain, yet 77% remain hesitant to allow AI to make operational decisions without human approval. The next phase of AI is not simply more automation, but AI that enterprises can understand, trust and govern. Riverbed is helping customers close these gaps with unified observability, trusted data, AI governance and intelligent automation, accelerating toward zero-disruption autonomous IT operations."