AI is changing work faster than organisations are changing with it
AI is changing work faster than organisations are changing with it
The workplace debate around AI has spent a lot of time focused on one question: will it replace jobs?
While it’s understandable, it’s increasingly not the most useful question to ask. OpenAI CEO Sam Altman recently acknowledged that fewer entry-level white-collar jobs have been eliminated than he expected and said he no longer anticipates the kind of immediate “jobs apocalypse” some have predicted. In Australia, Prime Minister Anthony Albanese has similarly argued that the goal should be for AI to support and create good jobs, rather than simply replace them.
Neither view means workforce disruption is off the table. Roles will change, some forms of work will decline, and new capabilities will be required. But they reinforce an important distinction: AI can have a significant effect on employment without replacing entire jobs at once.
Our own research points in a similar direction. Job security was named as the biggest concern by just 7% of respondents, down from 25% in 2025. Meanwhile, upskilling existing employees is the most common workforce response to AI, ahead of redesigning roles or workflows and changing the skills organisations hire for.
The change is already happening inside jobs. Nearly half of the 1,505 business leaders and technology professionals surveyed for our latest research use AI daily, while 71% use it at least a few times a week. What was recently treated as an emerging workplace technology is quickly becoming part of the working day.
The bigger question for employers is therefore not simply whether AI will replace people, but what happens to roles, skills and workforce decisions as people incorporate it into how they research, draft, analyse, plan and solve problems.
Employers are beginning to reconsider what good performance looks like. Hiring managers increasingly need to distinguish between people who can access AI and those who can use it critically, safely and effectively. AI adoption may be accelerating, but many organisations are still catching up with what that means for their workforce.
AI skills are becoming a baseline workplace expectation
For many roles, AI literacy is moving in the same direction as broader digital literacy: from a specialist capability to a basic expectation.
That doesn’t mean ever employee needs to be an AI expert, but it does mean employers are increasingly likely to expect people to understand how AI can support their work, where its limitations sit and when its output needs to be challenged.
As Matthew Munson, Managing Director at Talent Sydney, explains:
“AI skills have shifted from a nice-to-have to a baseline expectation for the vast majority of roles we recruit. We’re actively advising candidates on how to keep building their AI capability and stay abreast of emerging trends.”
But there’s an important distinction to be made between familiarity and capability.
Only 36% of respondents describe themselves as using AI to produce stronger, more useful outputs. A much larger proportion primarily use it for quick answers, research or one-off tasks. This suggests that while AI access is becoming widespread, the pool of genuinely capable users remains much thinner.
For hiring managers, asking whether a candidate “uses AI” will therefore reveal very little. A more valuable conversation explores how they use it, what they use it for, how they validate the result and whether it has improved the quality of their work.
The real change is happening inside jobs
The strongest evidence of AI’s workplace impact is currently found at task level.
Eighty-five percent of respondents say AI has made them more efficient at work, and more than half report meaningful time savings, while another 32% have seen gains across smaller or more specific tasks.
While it matters, efficiency doesn’t automatically mean a role has become simpler, or that the work has disappeared.
More than one in four respondents say AI primarily helps them start tasks rather than finish them, while 18% report spending more time reviewing or correcting AI-generated work.
In practice, this redistributes effort rather than removes it, with less time spent facing a blank page, gathering basic information or producing an initial draft. More time may then be required to assess quality, test assumptions, identify errors and turn a generic output into something useful in context.
The capability employers value will change accordingly.
Employees may be expected to produce work faster, but the strongest performers won’t simply be those who generate the most output. They’ll be the people who know which work should be accelerated, which decisions still require deeper thinking, and how to recognise when an AI-assisted answer is incomplete or wrong.
Human judgement is becoming increasingly more valuable
Loss of human judgement was the most common concern among survey respondents, ahead of data security, over-reliance and job security. At the same time, 57% identified relying on AI outputs without checking them as one of the highest-risk behaviours organisations now face.
This has clear implications for recruitment and assessing candidates.
A candidate can use AI to help prepare an application, respond to a technical question or complete an initial task. Employers therefore need stronger ways to understand the capability behind the output.
That could mean placing greater emphasis on work samples, follow-up questions and role-relevant scenarios. It may involve asking candidates to explain how they reached a conclusion, what they would verify and where they would avoid using AI altogether.
The objective shouldn’t be to catch people using AI. In many roles, its use will be both expected and beneficial, and the objective is to establish whether the person can apply it with sound judgement.
AI can make an average answer appear polished, but it can’t consistently replace the technical depth, commercial context or decision-making that determines whether the answer is useful.
Workforce planning is beginning to respond
Organisations are starting to adjust their workforce strategies, although the shift remains uneven.
Upskilling existing employees is currently the most common response, cited by 44% of respondents. Thirty-seven percent are reducing manual or repetitive work, while 22% are redesigning roles or workflows and the same proportion are hiring AI specialists. However, 26% say AI has not yet changed their workforce planning at all.
The emphasis on upskilling is sensible. The people already inside an organisation understand its customers, systems, processes and operating context, and that knowledge becomes especially valuable when applying AI to real work.
However, training employees to use new tools is only one part of the equation.
Employers also need to consider whether roles have been designed around outdated ways of working. If AI removes part of a task, what should replace it? If it increases output, who is responsible for checking quality? If a capability is now required across a function, should it remain concentrated in a specialist team?
These are workforce design questions, not simply technology questions.
What AI-capable talent looks like
There will not be one universal definition of an AI-capable employee. The required capability will vary by profession, seniority and the risks attached to the work.
However, the most valuable people are likely to combine several qualities:
- Relevant technical or professional expertise
- Confidence using AI in context
- Awareness of data and risk
- The ability to validate outputs
- The judgement to know when human intervention matters
As Thomas Mackenzie, Director of Client Services at Scale by Avec, puts it:
“AI is changing what great looks like, and the real advantage sits with the organisations that back people who can combine AI with the technical expertise and commercial judgement to deliver better outcomes, not just faster outputs.”
For employers, this means looking beyond generic claims of AI proficiency. For candidates and contractors, it means being prepared to demonstrate how AI has improved a genuine piece of work, not simply listing a tool on a CV.
The workforce can’t and won’t wait for a definitive answer on how AI will reshape every role, so the organisations best placed to respond will be those that treat AI capability as a workforce issue over just a technology one. That means defining what effective use looks like, redesigning work where necessary and backing people who can combine AI with the expertise, context and judgement it cannot provide on its own.
Our latest research reveals how more than 1,500 business leaders and technology professionals across Australia and New Zealand are navigating that shift. From everyday adoption and productivity gains to skills, trust and the future workforce, access the full report to see what employers should be preparing for next.