Every small business owner knows the difference that any amount of extra time would make to their week. More customers served. Better processes. More time developing people. Or perhaps getting home before dinner for once. 

AI appears to be helping us find this time. RMIT Online’s research with Deloitte Access Economics found workers save nine hours each week through AI use. Across a standard working year, that adds up to around 11 working weeks, or 2.5 months.

For a small business owner, that sounds like a huge win. Almost half of employees are using the time to complete more tasks. But providing an AI subscription and leaving employees to teach themselves to use the technology can introduce and increase operational, legal and reputational exposure risks that you might not be placed to absorb.

As AI becomes the top business risk concern in Australia, the stakes of balancing the benefits with the risks of their AI investment are about to become much higher if employees continue learning it through trial and error. What businesses should be leaning on instead is the opportunity gained from the use of AI, and what businesses choose to do with that time.

The real obstacle is (a lack of) judgement

As AI becomes part of everyday work, every unchecked output becomes another opportunity for an error to reach a customer, influence a decision or create rework. We’re already seeing this in practice as 95% of organisations in ANZ have experienced an AI-related incident, with around 77% reporting direct financial losses averaging $800,000 per company. 

To reduce the risk, the focus must shift from speed and technical AI skill development to judgement and critical thinking. At the moment, workers are roughly twice as likely to be advanced in technical AI skills as in judgement-based skills. What’s important to distinguish is that using AI quickly is not the same as using it well. 

Technical capability is easy to see. Employees learn to write better prompts, generate content faster and automate repetitive tasks. Judgement is less visible, but much more valuable. It means knowing when to trust an output, when to challenge it, and when AI should not be used at all. Accuracy, bias, confidentiality, legal boundaries; these are all things that must be assessed with judgement skills to make sure the benefits of AI aren’t erased by a mistake. 

While there is no overnight solution to make employees critical of AI-generated output, the small businesses that stand out will be those who invest in their people, ongoingly — and how they learn. 

How people learn matters

Given AI is an emerging technology, many employees started learning on the go to keep up. Nearly half of workers identify as self-guided learners, meaning they build their skills through trial and error. 

While experimentation can build technical fluency quickly, this unstructured way of learning is less reliable for developing judgement because employees can’t identify risks, they do not yet know exist. Offering training might be difficult to validate for a small business working with fewer resources, but structured upskilling doesn’t need to involve expensive programs. Instead, focusing on role-specific courses or modules can move people beyond basic prompting. For example, sales teams might practice validating AI-generated customer insights, while marketing staff focus on copyright, privacy and disclosure, and finance on checking calculations and protecting commercial data.

The next step is working with employees ongoingly to refine their AI skills alongside their judgement.

Beginners can complete simple tasks but often accept outputs without verification. Intermediate users refine prompts, move between tools and recognise when outputs need checking. Advanced users build verification into their workflow and judge whether AI is the right solution in the first place. This progression matters because most workers use AI, but just half remain beginners, and only 7% are advanced. 

The research doesn’t show that advanced users automatically save more hours. It shows stronger literacy enables genuine process improvement, better decisions, and reduced risk. 

Stronger AI capability =/ more work

We know AI saves 2.5 months of work, meaning it is up to businesses to decide how the reclaimed time will be used. When thinking about sustainable growth, the goal shouldn’t be nine hours of more activity a week — especially as workers with advanced AI literacy are twice as likely as beginners to feel the pressure to work faster.

Instead, it is about rethinking ways of working to make the whole business more efficient while preventing people from burning out. The extra time can be used for more work when it makes sense, but it can also be used to free up time for for strategic thinking, improving processes and further learning. 

Ultimately, small businesses that pair technical fluency with judgement, structured learning and realistic expectations will capture the time dividend without turning efficiency into a new source of risk.

AI has given us the gift of time; the opportunity is to decide how and where it’s used.