Artificial intelligence is changing how Australian businesses recruit, manage and make decisions about employees. But as the technology advances, are workplace policies and management practices keeping pace?
Artificial intelligence (AI) is no longer something businesses are simply considering for the future. It is already being used to screen job applications, prepare performance reviews, draft employment correspondence, summarise meetings and assist with workplace investigations.
For small and medium-sized businesses, the attraction is understandable. AI offers opportunities to reduce administration, improve consistency and free up valuable management time.
However, there is a growing disconnect between how quickly organisations are adopting these tools and how effectively they are managing the associated risks.
The fundamental issue is not whether employers should use AI. It is whether they understand the technology well enough to use it responsibly.
And recent developments in Australian employment and privacy regulation suggest this is becoming increasingly important.
The Fair Work Commission has introduced new AI requirements
On 20 October 2026, new requirements governing the use of generative artificial intelligence in Fair Work Commission proceedings came into effect.
Under the Commission's guidance, parties using generative AI to prepare documents must disclose its use, explain how it was used and verify that the information is accurate and relevant.
Additional requirements apply to witness statements and declarations, which must reflect the individual's own knowledge and words.
These requirements relate specifically to Commission proceedings. They do not establish a general legal obligation to disclose every instance of workplace AI use.
However, they reinforce an important principle for employers: using AI does not transfer responsibility for the accuracy, integrity or fairness of a decision.
Consider a manager who uses AI to prepare a formal warning letter following an employee misconduct investigation.
The document might look professional, reference relevant employment principles and present a convincing explanation for disciplinary action.
But what happens if it misrepresents the evidence, introduces allegations that were never put to the employee or relies on an incorrect interpretation of employment law?
The employer remains responsible for the process and the resulting decision.
A professionally written document is not necessarily a procedurally fair one.
AI in recruitment: efficiency versus fairness
Recruitment is one of the clearest examples of where AI can deliver substantial benefits while introducing risks that employers may not immediately recognise.
AI-enabled recruitment platforms can screen applications, compare candidates against selection criteria and assist with shortlisting.
Used appropriately, these tools can make recruitment more efficient.
The concern arises when employers begin treating AI-generated assessments as objective conclusions rather than information requiring verification.
For example, an automated screening process might rank candidates according to qualifications, career history or particular wording within their CVs.
But does the system recognise transferable skills? Can it appropriately assess non-traditional career paths? Could its recommendations inadvertently disadvantage particular groups?
The Office of the Australian Information Commissioner has specifically identified the potential for inaccurate or biased outcomes when AI is used in recruitment.
Employers should therefore be asking whether their recruitment systems are appropriately tested, whether recommendations can be explained and whether meaningful human oversight remains part of the selection process.
Ultimately, hiring decisions should be supported by evidence against genuine role requirements, not simply an algorithmic ranking.
New privacy obligations are approaching
Another significant development arrives on 10 December 2026.
Changes to the Privacy Act 1988 introduce additional transparency requirements for organisations covered by the Australian Privacy Principles when computer programs are used in certain decisions that could significantly affect individuals' rights or interests.
Where the statutory conditions are met, an organisation's privacy policy must describe the relevant categories of decisions and personal information used.
Importantly, these requirements are not limited to tools specifically marketed as artificial intelligence. They may also capture other automated systems.
Depending on how a system operates, this could be relevant to recruitment screening, employee assessments or other employment-related decisions.
Not every employer or automated process will fall within the requirements. Applicability depends on the organisation's obligations under the Privacy Act and the nature of the decision-making process.
Nevertheless, businesses should now be reviewing what technology they use, what information those systems process and whether their existing privacy documentation adequately reflects their practices.
Workplace investigations: where the risks become particularly serious
AI can assist with administrative aspects of workplace investigations, including organising information, preparing interview structures and identifying themes within documents.
However, employers need to distinguish between using technology to support an investigation and allowing technology to determine its findings.
Workplace investigations frequently involve conflicting accounts, sensitive personal information and difficult assessments of credibility.
These matters require careful consideration of the available evidence, the standard of proof, procedural fairness and the particular circumstances of each allegation.
AI-generated summaries may omit context, misunderstand statements or produce conclusions that are not adequately supported by the evidence.
There is also the question of confidentiality.
Uploading witness statements, employee records or sensitive allegations into an unapproved AI platform may create privacy and information security risks.
Employers must understand how information is stored, processed and potentially disclosed before introducing these tools into sensitive HR processes.
The principle should be straightforward: AI may assist the investigator, but it should not replace the investigator's independent judgement.
The overlooked risk: managers using AI without guidance
Perhaps the most immediate risk for many businesses is not sophisticated automated decision-making software.
It is individual managers using readily available AI tools without clear organisational guidance.
A supervisor might upload an employee's performance history to generate a performance improvement plan.
Another might ask an AI chatbot whether particular conduct amounts to serious misconduct.
A third might use AI to draft an outcome letter following a workplace complaint.
These activities may appear relatively harmless, particularly when the intention is simply to save time.
However, without appropriate safeguards, managers may inadvertently disclose confidential information, rely on incorrect advice or generate documentation that does not accurately reflect the circumstances.
A generic AI-generated performance improvement plan, for example, might contain reasonable-looking objectives but fail to identify the actual performance deficiencies, appropriate support or realistic improvement timeframes.
The result could be a process that appears structured but is fundamentally flawed.
Employers should not assume that managers understand these risks simply because they are comfortable using the technology.
Training, clear boundaries and appropriate approval processes are essential.
What does this mean for your business?
There is no need for employers to abandon AI or introduce unnecessarily complicated governance arrangements.
The objective should be to establish practical safeguards that reflect the size, complexity and risk profile of the organisation.
We recommend starting with six questions:
Do we know where AI is being used? Identify the tools being used across recruitment, HR administration, performance management and employee relations.
Do we have an AI usage policy? Establish clear expectations about approved tools, permitted activities, confidentiality and information security.
Are managers appropriately trained? Ensure employees understand that AI-generated material must be reviewed and verified before being relied upon.
Who remains accountable for decisions? Maintain meaningful human oversight, particularly where decisions affect employment opportunities, performance, disciplinary outcomes or termination.
Are our privacy arrangements appropriate? Review how employee and candidate information is collected, stored and processed, including whether the December 2026 transparency requirements apply.
Can we explain and defend our decisions? Ensure employment decisions are supported by reliable evidence, appropriate documentation and procedurally fair processes.
These measures are not intended to create unnecessary administration. They are about reducing risk while allowing businesses to benefit from the technology.
Finding the right balance
Artificial intelligence will continue to change the way businesses manage their people.
Organisations that use it effectively may benefit from faster processes, better access to information and more efficient administrative support.
But efficiency should never come at the expense of fairness, confidentiality or sound management judgement.
For employers, the challenge is to move beyond simply allowing AI to be used and develop a clear understanding of where it adds value, where it creates risk and what safeguards are required.
The businesses that manage this well will not necessarily be those with the most sophisticated technology.
They will be the ones that combine technology with capable managers, clear policies and defensible employment practices.
How Winnchester Consulting can help
At Winnchester Consulting, we help businesses navigate changing workplace requirements through practical, commercially considered HR advice.
Whether you need to review existing HR policies, strengthen recruitment processes, improve manager capability or ensure workplace investigations and performance management decisions remain fair and defensible, we can help.
AI may be changing how workplace decisions are prepared, but employers remain responsible for how those decisions are made.
Contact us to discuss how we can help your business establish practical, effective HR processes.
