For New Zealand businesses, the next stage of AI adoption is likely to be less about getting people to try AI and more about helping them use it confidently, productively and responsibly.
If your employees are already using ChatGPT, Microsoft Copilot or other AI tools, you might reasonably ask whether they really need AI training.
The short answer is yes.
Using an AI tool and knowing how to use it well are two different things. An employee might be able to ask AI to write an email, summarise a document or brainstorm ideas, but that does not necessarily mean they understand how to get reliable results, protect company information, check what AI produces or use it effectively as part of their job.
For New Zealand businesses, the next stage of AI adoption is likely to be less about getting people to try AI and more about helping them use it confidently, productively and responsibly.
Using AI isn't the same as being skilled at using AI
Most people don't need a computer course to start using Copilot or ChatGPT. The barrier to entry is low and the tools are designed to be intuitive.
That's part of the appeal. But it can also create a false sense of capability.
Someone may know how to type a question into Copilot but not know how to give it useful context, refine the response or recognise when the answer needs checking. They may not understand what information should never be entered into a public AI tool. They may also spend more time correcting a poor AI-generated response than they would have spent doing the task themselves.
Good AI skills go beyond knowing which button to press. They include understanding what AI is good at, where it can go wrong and how to incorporate it into everyday work.
What should employees actually learn?
AI training doesn't need to turn everyone into a technical expert. For most workplaces, there are some practical capabilities that matter more.
1. Understanding what AI can and can't do
Employees need a basic understanding of how generative AI works and why it sometimes produces convincing but incorrect information.
This helps people use AI with the right level of confidence. AI can be excellent at generating ideas, summarising information, drafting content and helping with routine tasks. It should not automatically be treated as an authoritative source.
2. Writing better prompts
The quality of an AI response is heavily influenced by the quality of the instruction it receives.
Employees can get significantly better results by learning how to provide context, explain the desired outcome, specify an audience and give examples. This doesn't mean everyone needs to become a "prompt engineer". It means learning how to communicate clearly with an AI tool.
3. Checking and improving AI output
AI-generated content still needs human judgement. Employees should know how to check facts, identify assumptions, spot missing information and improve the result rather than simply copying and pasting it.
This is particularly important when AI is being used for customer communications, analysis, reports, decision-making or other work where accuracy matters.
4. Using AI safely
AI training should also cover privacy, confidentiality, copyright, security and organisational policies. Employees need to know what information they can put into an AI tool, what they shouldn't, and when they need to use an organisation-approved application.
This becomes particularly relevant as AI moves from something employees experiment with individually to something organisations use across teams and business processes.
5. Applying AI to real work
Perhaps the most valuable part of AI training is showing people how it can be applied to the work they actually do.
- For an administrator, that might mean drafting correspondence, summarising meeting notes or working more efficiently with spreadsheets.
- For a manager, it could mean preparing reports, analysing information or creating first drafts of communications.
- For a project team, it might involve turning unstructured information into a useful starting point for planning.
The aim isn't to use AI for the sake of it. It's to identify where it can remove unnecessary effort or improve the quality of work.
Should everyone receive the same AI training?
Not necessarily. There are some fundamentals everyone should understand, including responsible use, privacy and how to evaluate AI output. Beyond that, training can become more useful when it reflects the different jobs people actually do.
A finance team, project team, customer service team and marketing team are likely to have quite different opportunities to use AI.
Role-based training can help employees move from "I know what Copilot is" to "I know where this can genuinely help me in my job."
AI confidence matters too
There is another reason training matters: confidence. Some employees will enthusiastically experiment with every new tool. Others will be cautious, unsure where to start or concerned that AI will make their role less relevant.
Neither response is necessarily a problem. But organisations need a workforce that can make informed decisions about technology rather than leaving AI capability to the most confident people in the room.
Practical training gives people a common foundation. It also creates a shared understanding of what good AI use looks like within the organisation.
The goal isn't more AI. It's better use of AI.
The question for businesses is no longer simply whether employees are using AI. They probably are. The more useful question is whether they are using it well.
AI skills training can help employees understand the technology, use it more effectively, recognise its limitations and apply it to real workplace tasks.
That makes AI training less about teaching people another piece of software and more about building the capability to work effectively in a changing workplace.
For New Zealand organisations, that is likely to become an increasingly important part of workplace technology training.
In short: AI adoption might happen organically. AI capability usually doesn't.