A recent report by Deloitte AI Institute found that a large number of employees are using generative AI in the workplace… without telling their bosses.
It is not surprising that the use of generative AI by the mainstream population is increasing. The release of ChatGPT in November 2022 opened the floodgates for competitors and eased access to these types of tools.
Social media has also helped drive adoption, with TikTok in particular providing content showing people how to use generative AI to automate tasks for them – from research, writing to even coding.
So it’s no surprise that workers are beginning to experiment with and implement generative AI into their own workdays.
The problem is that many workers are not exposed to the use of ChatGPT and other generative AI tools.
According to Deloitte’s Gen AI survey, 32% of the 2000 employees surveyed are using some form of Gen AI for work tasks. Two-thirds of these respondents (ie about 20% of all those surveyed) agreed that their managers might not know about it.
The same study found that only 9.5% of large Australian businesses and 1.4% of SMEs have officially adopted AI in their business.
But this is not the first time such statistics have been seen. Back in March, a survey conducted by Fishbowl revealed that 68% of respondents were using Gen AI for work without informing their organization.
Why this is a problem for businesses
There’s been a lot of talk about how general AI can be used to cut down on administrative time and allow people to focus on more high-level tasks – this can also be seen as a way for organizations to sell into AI. It can then use it to replace human jobs.
With this fear spreading, it’s not surprising that some workers don’t want to put themselves out there to automate part of their work.
But regardless of motivation, it’s a problem.
According to the same report, 70% of Australian businesses have yet to break ground in preparation for general AI implementation. In fact, Australia ranks second out of 14 leading economies when it comes to general AI deployment.
Combining a lack of corporate readiness with employees potentially feeding company data into public chatbots and large language models (LLMs) is a recipe for disaster.
We saw this in May when Samsung banned employees from using ChatGPT after uploading sensitive internal source code to the chatbot.
The current confused approach to generative AI by many Australian businesses is extremely dangerous. And that’s before you even consider what other organizations in your supply chain might be doing in this space, when there’s not much oversight.
Arguments for responsible AI adoption in businesses
As we enter this nascent phase of generative AI adoption, businesses need to be open-minded and cautious. It’s not likely to go away anytime soon, which means that if you’re considering using generative AI, or suspect your workforce will, it’s worth implementing an AI strategy.
Because it is not that easy to bring chatbots to automate tasks for you. Potential security risks, bill shock (despite the popular rhetoric it’s not all cost-savings), biases and AI illusions are still very real.
And while there are model standards and ethical practices to help guide organizations, we are years behind in terms of required legislation.
In these early years, when most people are still learning and making mistakes, if AI is going to be implemented, businesses need to have open communication and transparency to ensure it is done responsibly.
And not just to protect the organization, but the people who run it every day.
This story has been updated to more clearly express the percentage of respondents using Gen AI in the workplace.