Artificial intelligence (AI)

AI means computer programs that do things which normally need human intelligence.

Different types of AI can:

• look for patterns in data and use them to make predictions or recommendations
• communicate using natural human language
• create new content like text, images and computer code. This is called generative AI.

Used properly, AI can help you make better decisions, improve productivity and connect with customers.

Identify the problems you want AI to solve

Don’t use AI just for the sake of it. Make a list of the business problems or goals AI could help with.

You might like to refer to the goals and objectives in your:

• business plan
• marketing plan
• business goals
• digital strategy.

Choose the right solution

Once you’ve identified how AI can help your business, research the best solutions for your needs.

A good place to start is by looking at what you already have. Many of the digital tools you currently use in your business might have AI features. For example:

• word processing and spreadsheet software
• customer relationship management (CRM) systems
• marketing automation, email marketing and social media software
• graphic design tools.

If you need something else, you can choose between an off-the-shelf product or a custom solution.

Off-the-shelf tools

Off-the-shelf AI tools tend to be:

• cheaper
• quicker and easier to implement
• reliable, because they’ve been tested by many users.

Many off-the-shelf AI tools offer free trials or provide a basic version for free. This means you can experiment with AI without paying a lot of money.

Custom solutions

Custom AI solutions can be tailored to solve your specific business problems. They are usually more flexible and scalable than off-the-shelf software, so they can evolve along with your business.

However, custom solutions:

• tend to be more expensive than off-the-shelf products
• have ongoing support and maintenance costs
• need to be trained on high-quality, well organised data. 

Bring your staff on board

Before introducing AI, it’s important to teach your staff about its benefits and address any misconceptions.

Some ways to do this:

• Explain that you’re not introducing AI to replace employees. You’re using it to make their lives easier and enhance their existing skills.
• Get staff involved in researching and choosing AI tools.
• Train employees to use new tools effectively, safely and responsibly. This will help them see the benefits firsthand.
• Keep communicating with staff and answer their questions as they start using AI.

Find out how to manage change in your business.

Start out small and evolve

AI has huge potential for businesses of all sizes. But there’s no need to change everything overnight.

Start by trialling AI in 1 or 2 areas of your business. Try out a few different products to test their accuracy and see what works for you. This also lets you make sure you’re using AI safely and responsibly.

Once you’ve got a better sense of how your business could benefit from AI, you can start rolling it out in other areas.

AI is changing incredibly fast. In a few years, businesses will probably be using AI tools we haven’t even thought of yet. Stay across the latest developments to keep improving how you use AI in your business.

AI is a powerful technology, so it’s important that you use it responsibly.

Here are some things to consider to make sure your business uses AI safely, securely and ethically.

Decide who is accountable for AI

Your business is responsible for everything your AI tools do. To protect yourself and your customers, make someone high up in your business accountable for AI safety.

If have a lot of AI tools, you can also assign people to look after individual AI systems.

Create processes and guidelines

Set up processes and guidelines to explain how you use AI responsibly. You might like to capture these in an AI policy or strategy for your business.

Things you could include:

• things you do and don’t use AI for
• who is accountable for using AI responsibly
• how you identify and manage the risks of each new AI tool
• how you protect personal data and make sure you’re following privacy laws
• when you tell customers you’re using AI
• any training your staff need to use AI responsibly.

The National AI Centre has a detailed template you can use to create an AI policy. 

Understand the possible impacts of AI

AI tools aren’t perfect. Just like people, they can make mistakes, take things out of context or use information from unreliable sources. Sometimes they can even discriminate against minority groups.

This can have serious impacts on people. For example, if AI used in a hiring process has a bias problem, it could unfairly reject qualified candidates.

Before you start using a new AI tool, think about the potential harms it could cause to your business, your customers or any other people. You might like to ask your customers or clients about the impacts it could have on them.

Provide a way for people to challenge AI decisions or report problems with your AI systems. This includes a process for putting things right if your AI tool negatively affects someone.

Manage the risks

Different AI tools have different levels of risk. For example, a chatbot that talks directly to customers is riskier than using AI to draft social media posts.

Once you’ve identified the potential impacts and risks of an AI tool:

• assess how likely each risk is and how bad the consequences would be
• decide if the benefits of the tool outweigh the risks
• if they do, make a clear plan to manage the risks.

Find out how to manage risk in your business.

Test and monitor AI tools

It’s important to thoroughly test an AI tool before you use it in your business. This includes making sure it doesn’t have any privacy or cyber security issues.

Make sure:

• the developer or supplier has tested the tool properly
• the tool works the way you expect
• you understand what data the tool uses and how that data is stored
• the tool protects your and your customers’ information and will not compromise cyber security.

Once you’re using a tool, continue monitoring it for any changes in behaviour. Give staff a way to report problems so you can investigate and act on them.

You should also tell the tool’s developer about any issues you find.

Check your AI output for errors

AI tools need human oversight to work safely.

For example, generative AI tools can 'hallucinate' – present false or misleading information as facts. So it’s important to check AI-generated content and decisions before you use them or share them with others.

How much oversight a system needs depends on its risk level. But you should make sure the output of your AI tools:

• is accurate
• is relevant
• uses language that’s appropriate for your audience
• is unbiased and respects our diverse society.

You should also have a clear process to override or stop using an AI tool if something goes seriously wrong.

Be transparent with your customers

It’s a good idea to tell your customers how you use AI. Explain:

• what you use it for
• how you’re using it responsibly
• the benefits for them.

How transparent you need to be depends what you’re using AI for. For example, a customer probably doesn't care that you used AI to help you write a social media post. But they want to know when they are talking to an AI rather than a human for customer support.

Document your AI systems

Set up a register to keep track of your AI tools and how you’re using them. This should include general software tools with AI features.

For each tool, document:

• why and how you’re using the tool
• who in your business is responsible for it
• its features and limitations
• risks and potential impacts, and how you’re managing them
• any technical details needed to understand the system.

The National AI Centre has an AI register template you can use to get started