Andrew Douglas: We’ve hit the main topic, haven’t we?

Tanya Maithani: We did.

Andrew Douglas: Do you want to kick off the main topic?

Tanya Maithani: Yeah, this is a very interesting topic, Andrew, and we are facing it with a lot of companies which come to us, a lot of times have come to us.

Andrew Douglas: Let’s talk about what the main topic is. I think we should actually say that.

Tanya Maithani: Okay. Okay, before moving to the main topic, actually I’ll ask you a question. The question is basically, you see a job application, and you apply to that job application. You feel that you are the eligible candidate for this, but you get screened out even before you talk to someone. Who is responsible for it if that’s a bad decision? And do you have an answer?

Andrew Douglas: So you’re talking about my life really. Well, the answer is, it’s very hard to know, isn’t it? Because you don’t actually know the decision process and you usually don’t know the people you’re competing with.

Tanya Maithani: Correct.

Andrew Douglas: So at my level, I do know what are, if not the exact name of the applicants, I certainly know what is my capacity, and therefore I could, because of my age, I could bring in age-based discrimination, in Victoria, because of your look.

Tanya Maithani: It is.

Andrew Douglas: Here you’d say look, personal appearance is something.

Tanya Maithani: It’s true.

Andrew Douglas: Which is a discriminating tool.

Tanya Maithani: It is.

Andrew Douglas: No other jurisdiction, which is why I live in Victoria. I could have a bit of a shot at it, but for 99% of people going for a job, they have no reason, they have no understanding as to why they were screened out, and they could get a pro forma letter that says, “Look, thank you for your application. I’m afraid you didn’t meet the skills and experience required,” and you’d be no better off.

Tanya Maithani: Absolutely correct. Australia Merit Protection Commissioner said the same thing, that the company is liable for anything that the AI does if they are using an automated system. And that brings back to the main topic, which is.

Andrew Douglas: It does, well look at, we’re back.

Tanya Maithani: It did, we came here. So it is, is the AI death of HR or is it a new superpower? So we’ll be discussing that today and learn today.

Andrew Douglas: Yeah, I think the reason both you and I think this is an interesting topic is HR is destabilizing the non-selling parts of business. So it is looking at what is the cost of business, of the internal service delivery, and it is carving out a lot of that.

Tanya Maithani: It is.

Andrew Douglas: And so a number of HR practitioners have said to me, “Look, I’m really afraid that I won’t have a job.” And the answer to that’s really interesting, what is AI doing in a positive level? It provides us, it gives you a capacity to collect large amounts of information in simple tabulated format that’s easy and digestible. And if you have an appropriate system that sits around it, with a minimum amount of hallucinations, if it’s properly prompted. That’s fantastic. It also allows you to identify measurable behaviors, measurable performance, and to create a plan that sits around what good looks like, what isn’t good, and who’s performing against it.

This is stuff that was very, very hard to do. And it also allows you to build that system that collects that information. So if you’re not insured, if you don’t know what it is, you’re not educated, AI can do it for you. Now, they’re the really good things. And you can see at the HR partner level, a lot of the work that is done, which is not front facing, the documentary process, the analysis, the performance review, that’s going to be gone, if it’s done correctly, and it should be gone to do in this way. But the downsides are quite considerable as well. So at a lower level than HR, there will be roles that go.

Tanya Maithani: Correct.

Andrew Douglas: Some of those will be replaced by data-centric people who specialize in protecting the AI environment so it doesn’t breach privacy, I’ll come back to it, and that it is properly prompted. So it will create a technological job that sits aside from it, supporting HR. But the downsides are quite considerable. And maybe talk about the downsides. Let’s start off, and I know I’m prompting you, and you’re not AI, but let’s start off with the issue of, what are the risks without privacy around using AI as a tool for the management of people?

Tanya Maithani: No, of course. I think for HR it’s really important, because you hold a lot of personal information, sensitive information around people. So you really need to know what is personal information, what is sensitive information, what can go into AI, and what cannot go into AI, because there is an immense risk that sits in it, especially because Australia is developing its own privacy laws. There is a privacy act that is sitting, and there are a lot of changes that are happening in the privacy act.

So I think HR needs to be really, really careful around this. Other than that, and you’ll be surprised by it, I’m not sure if you will be surprised by it, but I think many people would be, that from 10th December 2026, there is a new rule coming that if you’re using AI in your privacy policy, you need to mention that, and it should be specifically there in the website. So that’s something that everybody needs to keep an eye on, because if you don’t meet that obligation, then the liability sits there.

Andrew Douglas: And can I, so this is a really interesting conversation, because the boundaries of AI are not completely understood. So if I put in people’s personal information into Claude, and I’m part of an environment where we’re using Claude, it’s not as easy as you think to put the administrative controls to prevent somebody else from finding out that information.

Tanya Maithani: Correct.

Andrew Douglas: So here we are talking about using it as a performance tool. We can use it as a demographic tool to understand, like our reporting conditions under sexual harassment, part of that is getting a greater demography of what’s inside our business. The very questions we’re asking Claude are capturing sensitive and personal information for which we don’t have adequate controls, or we should have adequate controls, and we need to actually ensure that Claude, or whatever other AI tool we use, has administrative controls to prevent the release of that information to anybody else.

Tanya Maithani: Correct.

Andrew Douglas: But it’s a big issue because AI is, by nature, porous, and always trying to grab and push out the information.

Tanya Maithani: It does that, yeah. Yeah.

Andrew Douglas: But the real crux of this is, from December 2026, you must have an AI policy, and you must have your privacy policy recognizing and relying upon that AI policy where it comes to private and sensitive information.

Tanya Maithani: That’s correct.

Andrew Douglas: Where are we going to end up with this is very simple. HR has always belonged in an operational area of the business. Too often it sits alone and it misunderstands the needs of operations. So it is going to be drawn towards operational support, and performance, and development.

AI is a fantastic teaching tool as well. But what it will mean is a rare opportunity for very senior HR people to embed themselves in the C-suite by being able to provide unique insights into what is the human capability that sits in the organization, what are the factors that are preventing peak performance and how that can be done.

So AI will damage the HR profession at the lower process level. We will see much more data management, much more technical management, but at the more senior level, it will enable higher performance, higher reward, and higher REM. So I think it’s fascinating where we’re going to go, and that will all happen in the next three to five years.

Tanya Maithani: It will, yeah.

Andrew Douglas: Why don’t we go to the case study? Do you want me to read or do you want to read?

Tanya Maithani: You give it a go.

Andrew Douglas: I’m going to read it.

Tanya Maithani: Yeah.

Andrew Douglas: Okay, Daryl had worked on the production floor for most of his adult life. At 47, he was tired. He had progressed to a leading hand role, and his crew loved him. He was tough and direct with them, and the screen keeps moving too quickly. He was tough and direct with them and he knew the ropes. The afternoon shift was tough. Most of the workers spoke English as a second language and had no relevant skills or experience. Staff turnover was high because the work was dirty and hard.

Daryl had recently applied for a salaried role in the production planning. The job description listed tasks that were within his experience and skillset. He completed the application and submitted it online through the HR portal. That evening, during the afternoon shift, Kathleen, the data analyst, questioned his numbers and waste levels. She was a young woman of 32 who had graduated university as a process engineer some six years earlier. To her, data spoke the truth.

Daryl’s shifts regularly had longer retool times and higher waste margins. The data may have been accurate, but it did not account for the fact that the afternoon shift workers were less skilled and experienced than those on the day shift. Daryl tried to explain, but Kathleen kept pointing to the charts. Eventually he told her to shove the tables up her, I think that’s a rude word. He was angry. How could she know so little about the human condition? The following day, Daryl was called to a meeting with Kathleen and Brad from HR.

It was a counselling discussion about his temper the night before. Daryl apologized for being gruff and explained that he was being treated unfairly over something he could not control, the quality of labor that was available. He also explained that all the small jobs were done on the afternoon shift, which meant constant retooling. So that’s changing the shape of what he was making and an increase in waste. He said that he tried to explain this, but Kathleen originally pointed to the charts and criticized him. He felt this was unreasonable.

Kathleen went red in the face and stormed out of the room. Brad shook his head and said, “Mate, when are you going to listen and reflect? You can’t speak to a girl like that.” Well, Daryl shook his head and said, “It’s not that she’s a girl, mate. It’s that she’s dumb, rude and doesn’t listen. Worse still, she’s wrong. Yet I’m the one who’s getting in trouble.” The following day, both Kathleen and Daryl were interviewed for the more senior production role.

Trev, the operations manager, raised the incident between Daryl and Kathleen with each of them when he interviewed them. Brad was present at both interviews. Daryl was called back after this interview. Trev patted him on the shoulder and said the job was his. He said Daryl had been right to stand up for what was right, and that the adjusted data, which took into account short production runs on the afternoon shifts, showed that he had outperformed the day shift. When Kathleen found out, she was devastated.

She held a bachelor of engineering with honors, had glowing reports, and had never spoken badly to anybody. She felt that old feeling that the boys club always won. Her mental health deteriorated and she presented a certificate of capacity. So we’re into the questions now, okay. Would Kathleen’s workers’ compensation claim succeed? It’s not a fair question for you. The answer is most definitely not in Victoria.

Tanya Maithani: That’s true.

Andrew Douglas:  The chances in New South Wales, after the legislation is amended under section 11, are not kind. So although it would be a bullshit claim, it would possibly be accepted in New South Wales. It would struggle in South Australia, struggle in Queensland, and Western Australia and Tassie, sort of somewhere, who knows what could happen? But certainly, in New South Wales, it’s a risk. Did the business discriminate against her or could she bring a successful general protections claim? What do you reckon?

Tanya Maithani: Not really. Daryl had the experience, which was much more than her, so I think it wouldn’t work.

Andrew Douglas: Yeah, she would have to be able to demonstrate, for the general protections claim, that she’d raised the workplace right in complaining about the behavior and she was terminated for that.

Tanya Maithani: Yeah.

Andrew Douglas: But what’s absolutely clear here is the decision maker was doing it because what Daryl said was right. Not his behavior was right, what said was right. His awareness and understanding of the situation was better than hers. Did Daryl’s conduct amount to a breach of safety law? If so, would it be sufficient to constitute serious misconduct considering that his language was mild compared with the language used day to day on the production floor? So there’s two answers here, and they sort of seem incompatible.

Yes, being gruff and rude is a psychological hazard and it’s a breach of section 21 in Victoria and like provisions throughout. In other words, what he did individually under section 25, again, slightly different provision elsewhere, is, did he exercise reasonable care to prevent harm to someone? Personally, no, so he would have a liability. Would never be prosecuted but have a liability. Organization, did the organization do everything reasonable and practical to prevent harm to a person who was working there and to monitor their health?

The answer is, organizationally, Daryl’s behavior would be attributed to the organization, so it would be a breach. Again, never litigated. But the next question is much more interesting, and that is maybe a safety breach, but is it serious misconduct under regulation 107 of the Fair Work Act, which is, did it create an imminent risk of harm? Was it in breach of a lawful and reasonable direction? So probably not in breach of a lawful or reasonable direction. There may have been a policy that prohibited it, but because there’s been condonation, there is other worse behavior done on a day-to-day basis. You remember we talked about the concept of waiver.

In workplace law we call waiver, that is waiver of a legal right. It’s called condonation, the doctrine of condonation. I cannot punish that which I permit. It would not be serious misconduct because it is conduct that is condoned. So if you can just remember, the Fair Work Act is a beneficial piece of legislation designed to protect workers, Daryl is safe. It would not be serious misconduct. It is misconduct, but it wouldn’t warrant termination of employment. So I think that’s where we’ve got to. I don’t think there’s much else to it, but I do want to just come back to this stuff.

When we’re talking about privacy, I don’t think most people who do workplace in HR and other parts understand how pervasive privacy law is at the moment and how much deeper it will dive into the way we work, behave, discuss things, and how quickly the use of technology, including AI, will create liabilities. So the takeaways from me is, do have an AI-based policy, and make sure your privacy policy and practices align with what is the current and future law which hits in December, because if you don’t, you’ll be in breach, and all workplace rights will come towards that breach.

Tanya Maithani: Correct.

Andrew Douglas: On top of that, clearly a breach of privacy law has two elements. One is, you can be charged for breaching privacy law. Remember that the commissioner is one person sitting on a fat lounger who does nothing, so it’s not really a great risk. Not quite so under the health provisions of Victoria, which is much more aggressive.

But there is also an emerging common law order, the Tort of Privacy which is happening, particularly with the publication of intimate details of people that could harm them. So I guess I just want you to understand, the world of privacy is much bigger than what you think, and it’s something we should be discussing now. Thanks for your time, see you later.

Tanya Maithani: Thank you.

Andrew Douglas: Cheers.