Independent Agency, In Marketing We Trust, partnered with the University of Sydney to research generative AI use in Australian businesses. It was found that mid-market companies were more concerned about generative AI quality than enterprise companies in Australia.
Participants interviewed by the University of Sydney were mapped based on the size and regulation level of their company and clustered into four main groups: lightly-regulated mid-market, lightly-regulated enterprise, highly-regulated mid-market and highly-regulated enterprise.
When it came to barriers to Australian companies implementing generative AI, five categories were identified within organisational reluctance: regulation, caution, detrimental to business, quality, and personal preference.
Regulation and caution categories were mentioned by all types of businesses. For regulation, businesses cited changes in regulation and non-existing legal frameworks as barriers to adoption. All businesses also showed caution, citing a careful approach regarding marketing, and fear of jumping on the bandwagon of technologies.
Surprisingly, mid-market companies were more worried about quality than enterprises, citing it as a main barrier to generative AI implementation. Both lightly-regulated and highly-regulated mid-market grouped businesses were worried about the wrong information and instructions shared by generative AI and acknowledged better quality by humans.
A lack of technical knowledge was also recorded in the research by In Marketing We Trust and the University of Sydney, with almost half of the interviewees not currently using generative AI. Forty-two percent of interviewees stated they did not use generative AI at all, while thirty-seven percent were experimenting with it, and just twenty-one percent were using it in their business.
An anonymous interviewee concerned about the accuracy risk of generative AI from a lightly-regulated mid-market Australian business stated:
“It spits out wrong information because of the garbage that is already in the system, you’re like you’re getting a false response and then you’re going to, I don’t know, deploy a media strategy which is not correct.”
