Surprisingly Useful AI Article Enhancements
Ai deals with the 4Ds: the dirty, the dull, the difficult and the dangerous. While these tasks may appear routine or peripheral, they often carry operational and security risk – making them ideal candidates for Ai support. Increasingly, Ai-powered digital operations are becoming the backbone of modern organisations, not because they replace human oversight, but because they augment it. From accelerating incident response to generating the code that underpins core systems, Ai agents have evolved from experimental tools to mission-critical infrastructure.
Yet as trust in Ai grows within Australian organisations, so too does recognition of the risks. Service outages, coding failures and security incidents involving Ai are no longer hypothetical, they’re happening.
Growing Trust In Ai During Crises
PagerDuty’s 2025 Ai Resilience Report, which surveyed more than 1,500 IT and business executives internationally, including 250 in Australia, revealed that 88 per cent of Australian executives trust Ai agents to act on their organisation’s behalf during a crisis. Many also acknowledge that their company would struggle to function without Ai, with 83 per cent saying it is essential to operate.
Organisations in Australia and around the world are entering a period of technological competition. Agentic Ai deployment is accelerating, with 88 per cent of Australian companies leveraging at least one Ai agent, with 22 per cent deploying five or more. As Ai models continue to mature and deliver better outcomes, executives are increasingly confident in offloading tasks to Ai. In fact, Australian leaders expect more than a third (35 per cent) of their department’s workload to be automated by 2026.
Existing Risks In Ai
Ai-generated code already underpins essential digital infrastructure. Today, 88 per cent of Australian companies use Ai to write, review or suggest code. However, only 32 per cent conduct formal, consistent checks for accuracy and reliability, meaning critical systems may often be built on code that lacks rigorous oversight.
When Ai-generated code fails, the consequences can range from service downtime to data breaches, often without a clear understanding of the root cause. Inconsistent quality controls leave businesses vulnerable to unexpected failures, which as Splunk’s 2025 Cost of Downtime Report highlights, can cost organisations around $200 million per year. While most Australian executives are confident in their ability to detect and mitigate Ai failures, real-world incidents suggest otherwise, with 85 per cent of companies surveyed already having experienced at least one Ai-related outage.
Resilience Planning Has Become A Vital Component Of Ai Strategy
Executives are increasingly aware of the risks that come with Ai adoption. 88 per cent believe their organisation needs better procedures to detect errors or failures in Ai tools. Even among companies that haven’t yet faced an outage are preparing for impact, with 78 per cent implementing protocols for Ai failure scenarios.
The PagerDuty survey highlights the new reality for operational resilience in Australia: organisations are becoming deeply reliant on Ai as adoption and trust accelerate and deliver measurable business results. However, as Ai becomes more embedded in critical operations, executives must prioritise reliability, governance and resilience to stay ahead of potential outages or crises.
Callum Eade is Vice President, APAC at PagerDuty.
Last Updated on October 16, 2025 by Callum Eade
