AI is arriving at work faster than many Australian businesses are ready for

Together, these charts show that Australia’s workplace AI story is not just about growth. Business AI research spending is rising quickly, but SME adoption remains uneven. Many firms are not yet broad users, many applications remain between awareness and use, and many expected benefits are still framed as possible rather than certain.

The bigger issue is readiness. If AI enters workplaces faster than businesses build skills, confidence and safeguards, the risk is not only slow adoption. It is uneven and uncertain adoption. For SMEs, the challenge is therefore not simply whether to use AI, but how to use it in ways that are trusted, useful and responsible.

I used OpenAI ChatGPT to support parts of this assignment, especially where I had difficulty with R, Plotly and CSS. Most of my use was for troubleshooting code, fixing page layout issues, adjusting spacing, colours and chart sizes, and making sure the visualisations stayed within the 600px width requirement. I also used ChatGPT to help improve grammar, wording choices and explanations.

The topic, data sources, chart decisions and final interpretations were selected by me. I found and checked the information from the Australian Bureau of Statistics and the National AI Centre myself. Because some National AI Centre dashboard data was presented through Power BI and was not downloadable as a CSV file, I manually entered the values into R and checked the entered values against the original dashboard sources.

ChatGPT was also used to help me understand some information and decide how to make the charts more readable for users. For example, in Chart 3, some percentage values were too small to fit clearly inside the blue bar segments, so I decided not to show those labels because they would make the design harder to read. I also tested different multivariate chart types, including heatmaps, but used bar-based designs where they could catch the attention of the audience better. I kept the heatmap for Chart 5 because I am trying to implement multiple charts for some users that can understand it as well. I am also open to feedback on how I should fix my visualization problems, if you are happy to share your though on this.

The visualisations use publicly available data from the Australian Bureau of Statistics and the National AI Centre. National AI Centre dashboard values were read using the all-time filter and manually entered into R for visualisation. Percentages are presented as response shares from the relevant dashboard categories, so some rows should be read as dashboard response shares rather than exact part-to-whole totals.