AI might help you reduce your workforce today. But are you also reducing your workforce’s capability tomorrow?
“We don’t need five junior employees anymore. AI can do most of their work.”
I suspect this is a conversation that will become increasingly common around executive tables.
And from a commercial perspective, you can understand why.
If technology allows a business to complete the same work with fewer employees, why wouldn’t it consider restructuring?
Reports that previously took hours can be prepared in minutes.
Research, administration, data analysis and document preparation can increasingly be completed with AI assistance.
For businesses under pressure to improve productivity and control costs, the attraction is obvious.
But there are two questions employers need to ask.
First, can we legally make those positions redundant?
Second, if we remove the junior positions today, where are our experienced employees going to come from tomorrow?
AI might be able to prepare the report, analyse the spreadsheet or draft the document.
However, that doesn’t necessarily mean the business no longer needs employees who understand how to do those things.
And under Australian employment law, replacing tasks with technology is not necessarily the same as establishing a genuine redundancy.
1. AI can replace tasks. But has it replaced the job?
Let’s imagine a business employs five junior analysts.
Their responsibilities include preparing reports, conducting research, reviewing data and providing administrative support to senior employees.
The business introduces AI software that significantly reduces the time required to perform those tasks.
The Management team decides it only needs three analysts.
Can the business make two positions redundant?
Potentially, yes.
Under section 389 of the Fair Work Act 2009 (Cth), a dismissal may be a case of genuine redundancy where the employer no longer requires the employee’s job to be performed by anyone because of changes in the operational requirements of the enterprise.
Importantly, a job can become redundant even where some of the employee’s duties continue to be performed.
For example, work previously undertaken by five employees might be redistributed among three employees, assisted by new technology.
The fact that some duties remain does not necessarily mean the redundant jobs still exist.
However, employers need to distinguish between eliminating a position and simply replacing one employee with another person performing substantially the same job.
The question is not whether AI has made the business more efficient.
It is whether the employer genuinely no longer requires the particular job to be performed by anyone.
AI may provide the commercial justification for a restructure. It does not automatically establish a genuine redundancy.
2. You still need to consult
One of the mistakes employers can make with restructures is treating consultation as something that happens after the decision has effectively been made.
Management identifies the efficiencies.
The new structure is approved.
The affected employees are told their positions are redundant.
Then HR is asked to manage the process.
That approach can create problems.
Section 389 requires compliance with any obligation in an applicable modern award or enterprise agreement to consult about the redundancy.
Depending on the applicable instrument, consultation requirements may involve notifying employees of proposed major workplace changes, providing relevant information, discussing the likely effects and considering measures to avoid or reduce adverse consequences.
Employers should also check whether any other contractual or statutory consultation obligations apply.
This is particularly relevant to AI-driven restructures.
An employer might be convinced that technology will remove the need for several positions.
However, have affected employees been given an opportunity to identify alternative ways of organising the work?
Could existing roles be redesigned?
Could employees be retrained?
Could the business achieve its productivity objectives without the same number of job losses?
Consultation does not necessarily mean employees get to veto a legitimate commercial decision.
But it should not be treated as a box-ticking exercise after management has already decided the outcome.
3. What if AI creates different jobs?
This is where I think AI-driven redundancies could become particularly interesting.
Technology may eliminate some tasks while creating new responsibilities.
For example, a business may no longer require employees to spend hours manually preparing reports.
But it may need employees who can review AI-generated reports, identify errors, interpret findings, manage data quality and advise management on what the information actually means.
The work has changed.
But has the need for the employee disappeared?
Not necessarily.
Section 389(2) provides that a dismissal is not a case of genuine redundancy if it would have been reasonable in all the circumstances to redeploy the employee within the employer’s enterprise or an associated entity.
That means employers should consider more than whether an employee’s existing position is disappearing.
They also need to consider whether reasonable redeployment opportunities exist.
In an AI-driven restructure, that might include newly created positions involving AI oversight, data analysis, quality assurance or other redesigned responsibilities.
Of course, not every employee will have the skills required for every new position.
And employers are not necessarily required to create entirely new jobs or provide unlimited retraining.
But an employee’s current job title should not be the end of the redeployment assessment.
Their existing skills, experience and capacity to undertake another role may also be relevant.
4. The High Court’s warning about redeployment
A particularly important decision for employers considering restructures is Helensburgh Coal Pty Ltd v Bartley [2025] HCA 29.
The case did not involve artificial intelligence.
It involved a mining employer that reduced its workforce following a downturn while continuing to use contractors to perform certain work.
A central issue was whether it would have been reasonable to redeploy employees who had been made redundant.
The High Court confirmed that the assessment of reasonable redeployment was not necessarily confined to existing vacant positions.
The Fair Work Commission could consider whether changes to existing staffing arrangements, including work being undertaken by contractors, would have made redeployment reasonable.
That does not mean employers must always displace contractors or reorganise their businesses to avoid redundancy.
The assessment depends on the circumstances.
But the decision is significant because it reinforces the need to consider redeployment practically, rather than simply asking whether there is an advertised vacancy.
Now consider how that principle might apply to an AI restructure.
An employer removes five junior positions but creates two new roles responsible for reviewing AI outputs and overseeing automated workflows.
The employer advertises those positions externally because it wants candidates with AI experience.
But did it properly assess whether the affected employees could perform those roles, including with reasonable training?
Were their transferable skills considered?
Were the new positions genuinely beyond their capabilities?
Before recruiting new employees to manage the technology, employers should consider whether existing employees could reasonably be redeployed into those roles.
That is not a conclusion that every AI-related position must be offered to an existing employee.
It is a question employers should be prepared to answer.
5. Can employers choose which employees stay?
Imagine AI allows a business to reduce its junior workforce from ten employees to six.
The business no longer requires four of the positions.
How does it decide which employees remain?
This is another area where employers need to be careful.
Selection decisions should be supported by legitimate, defensible criteria relevant to the future requirements of the business.
For example, an employer may consider relevant:
- skills,
- experience,
- performance,
- qualifications; and
- the capabilities required under the proposed structure.
However employers should be cautious about making assumptions.
Younger employees are not necessarily better at using AI.
Older employees are not necessarily less adaptable.
An employee who has taken parental leave should not automatically be disadvantaged because they have had fewer opportunities to develop new skills.
And an employee who previously raised workplace complaints should not be selected for redundancy because management regards them as difficult.
Depending on the circumstances, selection decisions may expose employers to discrimination or general protections claims.
A legitimate restructure does not provide protection for an unlawful reason behind selecting a particular employee.
Employers should document why positions are being removed, how employees are assessed and why particular employees are selected.
6. Are we removing tomorrow’s senior employees?
Putting the legal issues aside for a moment, there is a broader commercial problem that I think deserves more attention.
Where do experienced employees come from?
They don’t arrive with ten years of experience on their first day.
They develop it.
Junior employees learn by preparing reports, reviewing information, making mistakes, receiving feedback and watching how more experienced colleagues make decisions.
They gradually develop judgment.
They learn what good work looks like.
They learn how to identify problems before those problems reach a client, customer or executive.
AI can potentially accelerate parts of that process.
However it can also remove some of the experiences that traditionally helped employees develop those skills.
Imagine a graduate who has never had to prepare a report from scratch because AI has always produced the first draft.
Five years later, that employee is expected to review a complex report, identify inaccuracies and provide strategic advice.
How do they know what good looks like if they have never properly learned the fundamentals?
That is not an argument against AI.
Businesses should absolutely explore technology that improves productivity.
However, workforce planning cannot simply be about how many positions can be removed today.
Employers also need to consider what capabilities they will require in three, five or ten years.
If businesses reduce their investment in junior employees, they may eventually discover that the experienced employees they need are harder to find.
And potentially more expensive to recruit.
7. Before making the positions redundant, ask these questions
For employers considering an AI-driven restructure, I would focus on six practical questions:
- What has actually changed? Identify the tasks being automated, the operational efficiencies and why particular jobs are no longer required.
- Have we met our consultation obligations? Check applicable awards, enterprise agreements, contracts and other relevant requirements before implementing the restructure.
- What work still needs to be performed? Consider whether remaining duties can be redistributed or whether redesigned roles are required.
- Have we properly considered redeployment? Assess opportunities within the business and associated entities, including relevant transferable skills and reasonable training.
- How are affected employees being selected? Use defensible criteria and consider discrimination and general protections risks.
- What skills will the business need in the future? Consider whether today’s efficiency gains could undermine tomorrow’s workforce capability.
These questions will not prevent every redundancy.
Nor should they.
Businesses are entitled to make legitimate commercial decisions about how they operate, including decisions to introduce new technology and reduce staffing requirements.
Nevertheless, those decisions need to be implemented lawfully and with an understanding of their longer-term consequences.
Before making the decision
A question to ponder:
AI might help you reduce your workforce today. But are you also reducing your workforce’s capability tomorrow?
Need Assistance?
South Geldard Lawyers regularly assists employers, Business Leaders, CEOs, CFOs, HR teams and Executives with:
- Redundancy and restructure advice
- Workplace Bullying Claims
- Unfair Dismissal
- Sexual Harassment matters
- General Protections
- Psychosocial hazard management
- Manager and leadership training
- Performance Management
If you would like to discuss a difficult workplace decision or conflict, contact our Employment Law Team on jmamaril@southgeldard.com.au or 07 4936 9100.
