Companies are using AI to cut jobs without building the capability to use it in the longer term, new national research shows.
A report from Humanova shows 70% of Australian office workers use AI weekly but there is a growing capability gap among the workforce.
The National Workforce Report 2026 is based on a YouGov survey of 1039 workers in organisations with more than 25 staff; it shows seven out of 10 office workers use AI but only one out of eight of those workers are ‘fluent’ in its use.
It found most that respondents use AI as a shortcut; it classifies this majority as ‘dabblers’ or workers who learn to use AI occasionally but not for much more than summarising emails.
It says 58% of survey respondents fall into this category with AI fluency unevenly spread; senior leaders are five times more likely to be AI-fluent as the average worker and spend 3.1 times more working with it.
COMPANIES USING AI
This comes when employers are increasingly seeing AI as a chance to cut jobs: Humanova says its research found that 26% of senior leaders see AI as a chance to cut jobs, a percentage which rises to 32% in companies with more than 1000 staff.
It points out that organisations are tempted to use AI to cut staff, often in response to shareholder pressure, before they are capable of making it work.
“The question most business leaders are asking is, ‘how many of my people are using AI?’ That’s the wrong question,” says Dr Sean Gallagher, founder of Australian AI workforce research firm Humanova and the report’s author.
“Adoption is not the same thing as capability. You don’t
solve the second problem by celebrating the first.
“Organisations are rushing to redesign work around AI but the danger is that they cut before they build,” he says.
“If you make the organisation smaller before you make it smarter and more productive, AI becomes a cost-cutting story that brings risk rather than a transformation that delivers value.”
TWO-TIER WORKFORCE
The report indicates a structural gap is emerging in companies using AI; 25% of senior leaders are becoming AI-fluent compared to just 5% of the office workforce.
Senior leaders are also more than three times as likely to spend more time working alongside AI which allows them to develop real capability.
The report says this leads to what it calls ‘AI intuition’: the ability to know when to trust AI, which tasks to delegate and how to direct it effectively.
Most office workers, by contrast, operate in process-heavy, system-coupled environments where the friction of integrating AI into day-to-day work prevents them from engaging more deeply.
FIXING FLUENCY
The report warns that the arrival of AI agents (systems that can plan, act and execute multi-step tasks autonomously) will accelerate the divide, not close it.
The report says these ‘agents’ may reduce the need for some routine work for companies but will increase the need for those with AI intuition to work with it, capabilities which most workers have not yet been able to master.
“Agents will make your business faster and cheaper; they won’t make it smarter. Only your people can do that,” Dr Gallagher says.
“If organisations begin removing routine work before they have built higher-order AI capability, they risk removing the very on-ramp workers need to become fluent.
“You might not be able to avoid workforce change but you can avoid being unprepared for it,” Dr Gallagher says.
“The organisations that will thrive won’t necessarily be the ones that cut fastest. They’ll be the ones whose people have the capability to figure out what comes next.”
CLOSING THE GAP
The report highlights one of these companies that it says has done that: HR technology firm ELMO Software which turned to Humanova after its early internal AI adoption stalled beyond basic applications.
The report says that by treating AI as a cultural, not technological, change, it cut its sales preparation time from 10 hours to 15 minutes, automatically resolved more than 250 HR queries every month with 100% use of AI by staff.
ELMO president Joseph Lyons says: “The challenge is no longer the time it takes to implement a good idea but how quickly we can generate and test new ones. That’s where AI has the potential to deliver lasting gains.”
Dr Gallagher says: “Adoption numbers have told us what we wanted to hear for three years.
“The real question is which of your people are fluent, which are dabbling and what’s stopping the rest from progressing.
“Most leaders don’t know the answer. The ones who find out first will have a workforce the others can’t buy.”
