'AI use by stealth': Are organisations doing enough to protect themselves?

New report reveals AI's 'back-door' entry into organisations

'AI use by stealth': Are organisations doing enough to protect themselves?

A new report from the Australian Chamber of Commerce and Industry has revealed that businesses are aware of AI use “by stealth” but are not doing enough to minimise potential risks.

This refers to the informal introduction of AI in the workplace through existing software integration or employee use.

Businesses are “broadly aware of” the issue, “but are not taking the appropriate action to address the risks that come from this,” said the report.

Citing findings from Datacom, the report found that almost half of IT businesses in Australia do not have a staff policy on AI use.

Another half do not have staff awareness training on AI for employees, according to the findings.

Source: Australian Chamber of Commerce and Industry

AI's back-door entry

These gaps in policy come as the report pointed out that AI is stealthily making its way across organisations through software integration.

"AI is rapidly being embedded by software providers into existing software to make the overall user experience better," the report read.

"In most cases, we are hearing that users are not actively aware that they are using 'AI applications' and the use of AI is not a business decision as such but accepted through software updates introducing the new AI features."

According to the report, the disconnect between conscious AI use and the use of AI through existing applications can be an issue for employers who need to ensure data sovereignty.

"Organisations need to be just as aware of how their data and privacy are being assured and where it is being kept through the use of these applications just as much as if they are introducing new 'AI' applications," the report read.

Bring-your-own AI tools

Another potential back-door entry is through employees, who are utilising their own AI tools even without employers' consent.

"A worrying difference in risk perception and mitigation is emerging in that employees, who are generally signing up to the free ChatGPT or other generative AI tools, believe that AI is inexpensive, easy to use and implement versus the reality of business exploration which is the opposite," the report read.

An open approach to AI will be necessary to address these issues, according to the report. This approach includes training and trusting employees to use new tools at work.

AI's discreet back-door entry into organisations is just one of the topics discussed in ACCI's new report.

Andrew McKellar, CEO of ACCI, said the report seeks to provide insights and principles to guide businesses amid AI's global takeover.

"AI adoption has the power to fundamentally improve productivity and enhance flexibility to ways of working in the near term, with benefits to business and workers," McKellar said in a statement.