30/08/2026 9 minutos de leituraPor Rafael

Share:

Artificial intelligence has already become part of everyday work life — and a big chunk of that runs through emails.

Who hasn’t opened up ChatGPT or some other assistant to polish a message before hitting send?

The problem is that when this practice is misused, it can lead to some pretty serious consequences — and a recent case in Australia shed light on the issue in a way nobody expected.

Before we dive in, here’s a fun bit of trivia: the first email in history was sent in 1971, during an experiment to test whether two computers could exchange messages. The original message was simply QWERTYUIOP — the top row of keys on a keyboard. Since then, email has become one of the most widely used communication tools in the corporate world. And now, with the arrival of AI, those exchanges have become, let’s say, even less human.

In July 2026, the Fair Work Commission — Australia’s national workplace tribunal — made an unprecedented decision: it formally banned the use of AI in communications between an employee and their employer.

The trigger was a bullying complaint in the workplace, but what really caught the attention of the commissioner handling the case was the quality of the emails sent by the worker himself — long, repetitive, and packed with accusatory language, with everything pointing to them being generated by some kind of AI tool. 🤖

The case raises a question that goes well beyond Australia:

Where is the line between using AI as an ally in professional communication and letting it speak for you — with consequences you can’t even imagine?

Receive the best innovation content in your email.

All the news, tips, trends, and resources you're looking for, delivered to your inbox.

By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.

What Actually Happened in This Case

The episode started as a fairly typical workplace dispute — at least on the surface. An employee working as a caretaker at a residential complex in Queensland filed a formal bullying complaint against his employer. He alleged that the workplace repeatedly refused to clarify his duties, gave confusing and ever-changing instructions, along with other accusations related to how he was treated on the job.

Commissioner Sarah McKinnon, who handled the case, concluded that the worker had indeed been a victim of bullying. However, when she reviewed the emails exchanged between the parties, she noticed something unusual about the messages sent by the caretaker himself. According to her, the worker’s correspondence was unhelpful and appeared to have been prepared with the help of an artificial intelligence tool.

The messages were far too long. Repetitive in a way that human beings rarely are. The commissioner described the communication as lengthy, broad, full of generalizations, repetitive, and frequently couched in accusatory language. She went so far as to say it was no surprise the employer had, at some point, simply stopped responding. But that decision to disengage and shut down, as she put it, meant that many of the worker’s legitimate concerns were never meaningfully addressed — which, in her words, was bad for everyone involved. 📋

To prevent the bullying from continuing, the commissioner ordered that all future communication between the parties be precise and respectful. And that’s when the most striking part of the ruling came: there should be no use of artificial intelligence tools in preparing correspondence between the parties. She also recommended that messages be brief and to the point. The order went into effect immediately.

AI in Professional Communication: Ally or Trap?

Before making any moral judgment about what the worker did, it’s worth understanding the bigger picture. Using artificial intelligence to improve professional emails is a practice that has grown explosively in recent years. According to Mai Nguyen, a senior lecturer at Griffith University’s business school, it has become the norm for workers to use AI to sound more professional and use correct grammar.

But she raises an important warning: using the wrong prompt in an AI generator can exaggerate facts and make the tone of the text completely inappropriate. In other words, the use itself isn’t the problem. The problem starts when AI stops being a support tool and becomes the main voice of the communication.

When you ask an AI to write an email making a complaint, filing a grievance, or even reporting bullying, it will do so with impressive efficiency — but it hasn’t lived what you’ve lived. It doesn’t know exactly what the tone of the conversation was, the look on your manager’s face, or the weight of the silence in a meeting. It builds a coherent narrative based on the information you provide, but that narrative can be distorted, exaggerated, or even contradictory to other facts of the case. And when that message ends up in a tribunal, it needs to faithfully represent your experience — not a turbocharged version of it.

Professor Daniel Angus, a digital communication expert at the Queensland University of Technology, found the case particularly interesting. According to him, there is a natural appeal in wanting to use technology to make a text stronger or more persuasive. The danger is that someone can use it to overwhelm an opponent in a debate — especially in an adversarial setting like a workplace dispute. 😬

Should You Disclose That an Email Was Written by AI?

This is one of the thorniest discussions happening right now. Professor Angus argues that we need to normalize much more the practice of disclosing when AI was used — but he admits it’s a complicated topic. There are people who are strongly against the use of AI, and because of that, many prefer not to be transparent about when they’ve turned to these tools.

Legal researcher Guzyal Hill, from the University of Melbourne, points to a revealing fact: very few organizations have well-defined AI policies, although there is growing pressure for that to change. She notes that it’s virtually impossible to escape AI in modern society, especially when we’re talking about large language models. Since Australia doesn’t have a specific national law on AI, each organization needs to develop its own internal strategy and policies.

For Dr. Hill, who serves as Chair of AI Governance at the law firm HWLE Lawyers, there isn’t much of an issue with using AI to write emails — as long as it’s done skillfully, within the law, and in line with each company’s internal policies. The key, she says, is always in how the tool is used. And she raises a specific warning: its use should be discouraged when private and confidential information is being fed into the AI generator. 💡

Can You Tell If an AI Wrote the Email?

Here’s a curious revelation: although there are telltale signs when AI writes a text, humans are notoriously bad at spotting them, according to Professor Angus. He cites an example that has become almost a symbol of the AI era: the em dash (—), which he considers a true casualty of this technological moment, so closely associated it has become with machine-generated text.

Another giveaway of AI use is tone of voice. You might read something and think: well, that doesn’t sound like this person. It’s that strange feeling that the words don’t match the way the person actually communicates day to day.

Angus also raises a warning that sounds like it’s straight out of a sci-fi movie. He talks about the growing outsourcing of emails and other forms of communication to AI agents, which would put workplaces in a really weird and almost dystopian position: a bunch of AI agents talking to each other while the actual workers remain practically oblivious to what’s being said.

Professor Nguyen has a practical and revealing take on this. She reads up to 30 emails a day and suspects that about 80% of them are AI-generated — especially the longer ones, which she can easily predict came from some kind of automated tool. Her advice for anyone using the technology is simple: review twice to make sure the text is clear, accurate, and actually says what you intended to send. After all, at the end of the day, the email goes out with the sender’s name on it. 🎯

Tools we use daily

What This Case Changes — and What It Reveals

The Fair Work Commission decision is unprecedented, but it probably won’t be the last of its kind. As artificial intelligence becomes more present in workplaces, legal systems around the world will need to develop clear criteria for what’s acceptable when it comes to AI-mediated communication — especially in legal and employment contexts. The Australian case sets an important precedent: tribunals can and should question the origin of documents and messages presented as evidence, including emails.

But beyond the legal dimension, this episode reveals something about how people are approaching AI at work. There’s a growing tendency to use these tools not just to improve writing quality, but to amplify emotions, reinforce arguments, and in some cases, construct narratives that go beyond what actually happened. That’s concerning — not because AI is inherently malicious, but because it follows the instructions it receives, and when those instructions come loaded with frustration, anger, or a specific agenda, the result can be a message that looks fair and balanced on the surface but is actually an extremely biased piece of persuasion. In bullying cases, where subjective perception is already naturally complex, this can completely distort how the facts are read.

The takeaway here is that artificial intelligence is a powerful tool — and like any powerful tool, it demands responsibility from the person using it. Handing your voice over to an AI in a sensitive situation, whether it’s a bullying dispute, a workplace negotiation, or any other communication that could end up in a tribunal, is a real risk. Not because AI always gets it wrong, but because it isn’t you — and in moments like these, being you is exactly what the process requires.

Transparency in AI Use: A One-Way Street

One of the discussions this case ignites head-on is the need for transparency when AI is involved in producing emails and other professional documents. Some companies are already starting to create internal policies on the topic, requiring employees to flag when a text was generated or significantly edited by AI — especially in formal communications, HR records, and any material that could be used in internal or legal proceedings. This is a trend that should grow significantly in the coming years as more cases like the Australian one come to light.

From a tribunal perspective, the issue is even more delicate. How can a judge or arbitrator know if an email was written by a human or by AI? Today, that distinction is still possible in many cases — precisely because of the patterns the Australian commissioner identified: texts that are too long, overly uniform, with elaborate language that doesn’t match the person’s communication history. But as artificial intelligence models become more sophisticated, that detection will get harder and harder. That means the rules need to come before — not after — the problems multiply.

At the end of the day, the Fair Work Commission episode is a wake-up call arriving at a strategic moment. AI in the workplace is here to stay, and that’s great — it boosts productivity, improves writing quality, and democratizes access to clearer, more professional communication. But using these tools responsibly means understanding their limits, especially when the subject involves bullying disputes, tribunal proceedings, or any situation where the authenticity of the human voice isn’t just a detail — it’s the center of everything. 🎯

Picture of Rafael

Rafael

Operations

I transform internal processes into delivery machines — ensuring that every Viral Method client receives premium service and real results.

Fill out the form and our team will contact you within 24 hours.

Related publications

Google AI: March announcements in technology and artificial intelligence.

Google AI in March: an honest recap of what was (and wasn’t) announced, and why expectations differ between experts and

AI and ROI: Adopting solutions in the company without the hype.

Results-driven AI: companies demand real ROI, cut costs, boost productivity and improve service with practical solutions.

OpenAI Artificial Intelligence: Multimodal Models, Automation, and Unified Data

Weekly AI roundup: news, autonomous agents, open models, platforms, and their impact on marketing and product.

Receba o melhor conteúdo de inovação em seu e-mail

Todas as notícias, dicas, tendências e recursos que você procura entregues na sua caixa de entrada.

Ao assinar a newsletter, você concorda em receber comunicações da Método Viral. A gente se compromete a sempre proteger e respeitar sua privacidade.

Rafael

Online

Atendimento

Website Pricing Calculator

Find out how much the ideal website for your business costs

Website Pages

How many pages do you need?

Drag to select from 1 to 20 pages

In just 2 minutes, automatically find out how much a custom website for your business costs

More than 0+ companies have already calculated their quote

Fale com um consultor

Preencha o formulário e nossa equipe entrará em contato.