22 November 2025
In a world where the use of AI tools has increasingly become common, we’ve become increasingly comfortable with its use.
AI has seen a huge popularity boom in the past year. While the technology certainly isn’t a new one, the last 18 months or so have marked a real turning point in how people are adopting it. In the workplace, AI tools from chatbots, to well-known large language models (LLMs) like ChatGPT and Gemini are being used in some capacity.
Now the question employers are asking is: “What legal and compliance risks could there be in using AI?”
According to The EY Australian AI Workforce Blueprint, more than two-thirds of Australian office workers have used AI in the past month. While AI does have clear productivity and efficiency benefits, more and more employers are asking us about possible risks of using these tools.
The real risk is in how people are using the technology. We’ve seen plenty of news cropping up about people relying on AI without checking for accuracy or expert opinions, to disastrous results.
Let’s take a look at how AI impacts compliance in the workplace.
A common risk with using AI tools is the risk of hallucinations, that is, when Large Language Models (LLMs) like ChatGPT create false or misleading information.
When employees use LLMs and rely on output without cross-referencing sources or checking for accuracy, it can lead to less-than-desirable results. Take, for example, the growing number of legal briefs that were created with the help of AI that are found to have false or even made up citations.
The nature of hallucinations in AI tools means that this problem isn’t confined to the legal industry. Any employee that doesn’t fact-check their output and relies on AI is prone to committing mistakes that could put the business at risk.
The best way to mitigate the risk of hallucinations and misinformation when using AI tools is by ensuring that all content created with AI is reviewed by a human before it’s used. Many employees are now using AI to create reports, documents, even social media posts for company use, and these must always be checked for accuracy before being used or sent out.
Don’t rely on AI for legal, HR, or safety decisions. An easy example: many people rely on AI overviews when doing online searches. These AI overviews are good at finding relevant results…but can get information wrong at times. As we’ve said above, always check for accuracy, and remind your employees to do the same.
Almost every aspect of work has an AI tool for it nowadays, and the hiring process isn’t an exception.
Run a search for ‘AI recruiting software’ and you’ll be inundated with plenty of tools. Some companies are using AI-driven tools to:
For employers with no dedicated hiring or recruitment resource, these tools are a huge help. However, there are key compliance risks every business owner or employer should be aware of.
Australia has anti-discrimination laws where it is unlawful to discriminate on a basis of protected attributes. These include age, disability, sex, race, intersex status, gender identity and sexual orientation in certain areas of public life.
If an AI-powered tool excludes or favours candidates, even unintentionally, it can open up a business to risk from breaching the law.
As with everything that comes with using an AI tool, it’s always important to double-check any outcome they make. The tools can still misinterpret CVs or make incorrect assessments.
As with traditional recruitment efforts, AI-powered tools collecting information on candidates must comply with Australian privacy obligations. Businesses need to be clear about what data is being uploaded into these tools, where they’re being stored, and who can access it.
Is your team using AI tools for work? Consider having an AI policy in place. Chat with our team about a tailored AI policy for your business.
When it comes to using AI tools in recruitment, the key is to remember that human oversight is still needed at every stage. Don’t treat it as a decision-maker. Also remember to document your recruitment process and indicate when and how AI is used.
Who owns AI-generated work?
If your employees are using AI to create content like blog posts for your company website, or documents for clients, the question of intellectual property will inevitably arise.
Potential copyright issues might stem from content generated using AI tools. If an employee generates any kind of content that resembles existing copyrighted material, it could open up a business to copyright infringement issues.
Inputting confidential information, like client details and sensitive data or proprietary company information, into a public AI tool is also a huge risk. This information can be used to train the model, and also means the information is now in the public domain even if it isn’t made apparent to the user.
When using any AI tool, be sure to review its terms and conditions—especially if it’s a free one. Better yet, ensure that employees only use approved tools that have clear commercial-use rights.
Be clear with your team about what information can and cannot be entered into any AI tool.
As the use of artificial intelligence in the workplace becomes more and more common, leaders like yourself will need to clear expectations around how these tools can be used.
Your employees can’t be expected to know the proper etiquette for the use of AI tools on their own, and an internal AI use policy will set expectations and create clear boundaries about what is acceptable and unacceptable when it comes to AI usage.
A tailored AI policy should:
If you and your team are using AI in your workplace, now (actually) is the right time to have guidelines in place.
Our team can help you create a clear, practical AI policy tailored to your business and is compliant with Australian workplace requirements.
