Artificial intelligence is here to stay, but there’s still not much clarity around how small businesses should be using AI responsibly – or indeed, how small businesses are using AI at all.
A new Government-funded report by UTS (“In their words: perspectives and experiences of SMEs using AI”) has some of the answers.
It highlighted that small businesses are overwhelmingly using generative forms of AI, such as ChatGPT. The report’s small-business respondents also see AI as a free-for-all, unmoderated space. Small-business owners surveyed had largely not developed any benchmarks for responsible AI use, nor were they aware of AI government standards already developed by the Government.
“[I think Australian SME workers would value clear] rules and guidance, like checklists, on what they need to check to ensure they’re complying with relevant regulations,” said one respondent, who was an employee of a small business in the professional, scientific, or technical services sector.
Small-business owners are concerned about AI’s data risks
Many SMEs remain hesitant to go all in on AI due to concerns about data risks, including potential privacy issues related to the sharing of customer and private data with AI platforms.
“We just need to be mindful of not putting… confidential information into Chat GPT,” said one business leader surveyed.
The report also highlighted that many SMEs feel limited in their awareness and familiarity with AI, and would value greater education, training and support.
“Many of our staff, especially beyond the technical units, are tech laggards,” said Muhammad, a small-business leader in the information media and telecommunications sector. “AI will be forced upon them at some point. A lack of desire to understand benefits and risks means that at some point someone is going to use it badly, possibly leading to data breaches.”
Adam Berry, Deputy Director, UTS Human Technology Institute commented, “It is clear that Australian small and medium businesses are beginning to enthusiastically adopt generative AI, but in the process, they are confronting the real practical complexities of that adoption. They are seeing poor quality outputs, worry about sharing data, and are navigating a difficult and fast-moving space.
Berry added, “We can do more to provide simple, practical, useful support to those businesses so that they realise the promise of AI, without becoming exposed to its risks.”
The UTS Human Technology Institute, who authored the report, is currently working on the Safe Artificial intelligence Adoption Model (SAAM) project. SAAM aims to build simple online tools and practical resources to help Australian SMEs capitalise on the benefits of AI while minimising exposure to risks.
What are SMEs really using AI for?
Two thirds of the small-business leaders surveyed used AI primarily for content generation, such as social media, ad copy, blogs, formal reports, emails, and more. Some also referred to activities related to optimisation and productivity.
The report noted that none explicitly discussed commonly cited AI use cases like predictive analytics, sales forecasting, supply chain optimisation, recommender systems, and robotics.
When asked about AI risks, nearly half of the responses from SME leaders identified concerns around AI accuracy and identified the need for oversight and review practices. One business leader said she was worried about the risk of potential brand damage.
“[There is risk of] potential brand damage from spelling and grammar that doesn’t sound like us,” said Mai, who works in the professional, scientific, and technical services sector.
“[I worry about the] accuracy of the information provided – [it is not yet] 100 per cent, so have to double-check,” said Dwayne, who works in the same sector as Mai.
