For employers confronting the rapid rise of artificial intelligence, the key question is no longer whether employees are using it, but how they are using it.
Recent decisions by Australian courts and the Fair Work Commission show that workers have already lost jobs, harmed their careers and become embroiled in legal disputes after relying on AI-generated material for tasks ranging from legal research to workplace correspondence.
In one of Australia’s most striking AI-related legal errors, a South Australian solicitor told the Federal Circuit and Family Court that she had dismissed her paralegal after AI-assisted documents prepared for a family law appeal contained what the court described as ‘non-existent, inaccurate and misleading authorities’.
The court found the filing included AI ‘hallucinations’—inaccurate and misleading references to case law that had not been checked before submission.
It concluded that the use of AI had produced ‘erroneous citations’ and referred the lawyers involved to their respective legal regulators.
Another case centred on Marcus Wibmer, whose employment at Fujifilm Data Management Solutions unravelled after he took offence at a colleague’s Microsoft Teams message. What followed was a months-long workplace dispute involving escalating complaints and increasingly confrontational emails drafted with AI assistance.
Fair Work Deputy President Tony Slevin described Mr Wibmer’s reliance on artificial intelligence as ‘unfortunate’ and ‘counterproductive’, saying it ultimately ‘led to his demise’.
‘The communications it produced were dense, repetitive and often rambling,’ he said.
Employment law expert Mia Pantechis (pictured) said the cases were likely to be the first of many, predicting AI policies would soon become as common as social media policies
Research from Microsoft’s 2026 Work Trend Index indicates that many workplaces are struggling to keep pace with the rapid adoption of AI
Courts and tribunals are increasingly confronting the consequences of employees using AI
‘They were demanding and overbearing. They lacked context and perspective.
‘AI appears to have given Mr Wibmer a false sense of security that his communications, laden as they were with allegations of impropriety by his managers and demands for corrective action, were appropriate and acceptable in a workplace setting. They were not.’
AI tools such as ChatGPT, Gemini and Copilot are also featuring in an increasing number of disputes before the Fair Work Commission.
In a 2026 bullying case, Commissioner McKinnon criticised correspondence that appeared to have been prepared using AI, describing it as ‘lengthy, wide-ranging, replete with generalisations, repetitive, and often couched in accusatory language’.
Although the commission ultimately found that the worker had been bullied, it also noted that the AI-generated correspondence had contributed to the deterioration of relations between the parties. The decision highlights the growing scrutiny of how employees use artificial intelligence in workplace communications.
The cases emerging across Australia suggest that employers, courts and workplace regulators are becoming less focused on AI use itself and more concerned about its consequences—especially when employees rely on it to prepare legal submissions, manage disputes or communicate with colleagues and managers.
In each matter, the central problem was not the use of AI, but the failure to check, verify or accept responsibility for the material it generated.
Maurice Blackburn employment lawyer Mia Pantechis said these cases were likely to be the first of many, predicting that AI policies would soon be as commonplace as social media policies.
Fair Work Deputy President Tony Slevin said reliance on artificial intelligence was ‘unfortunate’ and ‘counterproductive’ in the case of one dismissed employee
DON’T MISS: Concertgoers Injured in Crowd Crush at OsamaSon’s NYC Show
‘This isn’t a new type of misconduct. It’s an old one with a new tool,’ she told the Daily Mail.
‘Workers have always been accountable for what they put their name to. AI just makes it faster and easier to get it wrong, and at scale.
‘We’ll see more of these cases, not because workers are careless, but because employers have handed them powerful tools, often without clear rules, training or guardrails.’
Ms Pantechis said responsibility did not rest with employees alone.
‘Employers who roll out AI, or expect staff to use it to work faster, without clear policies or training, can’t then act shocked when things go wrong. Accountability has to run both ways.’
She said the early disputes could mark a turning point in how workplaces manage the risks associated with artificial intelligence.
‘These early cases will be remembered as the moment workplaces were forced to work out who carries the risk when AI gets it wrong. Workers shouldn’t be left carrying all of the risk.’
Ms Pantechis also warned that AI could give employees misplaced confidence during workplace conflicts.
Centre for Future Work policy director Fiona Macdonald (pictured) said workplaces were still scrambling to establish rules governing artificial intelligence
‘When employees are stressed and in conflict at work, AI can be the perfect ‘yes man’ and provide a false sense of security about the communications being sent and threatened legal claims.
‘AI might sharpen your grievance, but it won’t tell you to sleep on it, and it won’t be the one who gets sacked if the communications or behaviours are unjustified.’
Research from Microsoft’s 2026 Work Trend Index suggests that many workplaces are struggling to keep up with the rapid adoption of AI.
Only 26 per cent of AI users said their leadership was clearly and consistently aligned on the technology. The report found that employees were often adopting AI faster than their employers.
Fiona Macdonald, policy director at the Centre for Future Work, said workplaces were still scrambling to develop rules around artificial intelligence.
‘The use of AI is speeding past any governance or systems that we have in place on how we use it and when we can use it,’ she told the Daily Mail.
‘A lot of workplaces don’t [have guidance], and they’re racing to catch up, racing to understand what they need to have, while employees are potentially thinking they’re doing the right thing and steaming ahead.’
Ms Pantechis said employees should remember that responsibility rests with them once they send a message.
‘You can’t blame the software,’ she said.
‘Never send anything you haven’t read, understood, and would be happy to defend in a disciplinary meeting. And if you’re in a dispute with your employer, talk to your union or a lawyer, not a chatbot.’