Gen AI is widely trusted but there are concerns
A survey of 60 chief financial officers (CFOs) of billion-dollar American enterprises reveals 97% of them trust generative artificial intelligence (gen AI) to manage risk. But almost 30% of them have concerns the technology’s capabilities and potential can be overrated.
Were a comparable survey undertaken in Australia, Macks Advisory believes the concern rate would be much higher, given what is already known locally about attitudes to AI.
Surveys here indicate less than half of us are comfortable with and trust the use of AI at work, and only a minority believes its benefits exceed its risks.
Furthermore, Australian attitudes are mirrored to some extent in the UK, Canada and France where it’s reported “fear and worry about AI are dominant emotions” – although proportions of these emotions relative to efficacy of the technology or relative to the job losses it could cause, aren’t specified.
However, our newsletter readers who use gen AI in their businesses might be interested to compare their experiences with the technology to what’s happening in the US.
The American perspective
Most CFOs say gen AI yields measurable value, and in sensitive roles its output is trustworthy. Thus, it is belief rather than figures that is turning the technology into an indispensable item in many a corporate toolkit.
A survey earlier this year shows the speed at which this is happening has accelerated dramatically.
The almost nine in 10 US CFOs surveyed who say they are seeing a “very positive” return on gen AI, are three times the number of those who expressed this view last year.
But it’s behind productivity gains where research further reveals an extraordinary level of trust among CFOs in the accuracy and reliability of the technology.
- There’s close to universal trust of gen AI in critical functions of business, 91% of surveyed US CFOs saying they have complete or a high level of trust in the technology across 10 core areas of business, 68% of them saying it’s “very important” in risk management, and 97% declaring they have high or complete trust in its output.
- The technology gets a 76% rating as “highly important” for strategic planning, although 98% of surveyed CFOs say they have strong trust in what it produces for this area.
- The same level of trust is accorded AI’s capability in financial reporting, which however gets only a 74% rating for effectiveness in this field.
So, obviously there are concerns
While trust clearly exceeds concerns, 29% of those who responded to the survey say gen AI may lack insight – and indeed this view seems to run parallel to the thinking of clients who have said things to Macks Advisory like this: “AI does a lot of wonderful stuff with what’s known, but it has no imagination.”
Furthermore 28% of CFOs flag risks of unauthorised access to their companies’ AI outputs, and 21% say they’re worried about bias than can be imbedded in the technology.
But if you’re running a business and considering what part AI might play in it, you can have confidence that more than two-thirds of people surveyed in the US have been prepared to dismiss the above concerns and have opted to employ the technology for the good things it’s been able to do for them.
However, they are doing this aware that to get the best out of AI, human intervention is necessary.
Almost three quarters of CFOs have told the survey team they’ve discovered they need staff involvement to generate new content with gen AI, and nearly two thirds say human review remains essential, even for low impact tasks like drafting employee feedback.
The bottom line here is that in adopting it, most CFOs are not simply betting on the efficacy of AI, they are betting on their ability to control it, in feeding it with trusted internal data and keeping its operation in context with a business’s function where oversight is possible.
Prime significance of intelligence data from the survey on gen AI’s use, is that return on investment (ROI), alone -- ROI is net profit divided by cost of investment x 100 -- doesn’t explain the rapid mainstreaming of AI among businesses.
This has also largely occurred because of a firming of the trust factor – built on data provenance, cautious implementation and human review.
Accordingly, confidence has been boosted to a scale that’s generated willingness to put AI at the centre of involvement in sensitive corporate decisions.
Australia beware!
Australian employers and employees have been put on notice by a global education company’s report that more than a quarter of the nation’s jobs are at high risk of being lost to automation and AI unless there is major upskilling in the workforce.
Commissioned by education provider and industry analyser Pearson, the report says Australia is facing a $10b economic black hole of lost earnings because of automation and its flow-on effects.
Industries most in danger of suffering from this forecast 26% job loss, are construction (nine percent) and manufacturing (six per cent). Automation technologies themselves – such as robotic processes, large language model chat boxes, AI models and autonomous mobile robotics, are all threatened by employers’ need of employees for entirely new, or much changed, jobs.
Jobs in retail, finance, and the media are also under threat from evolving gen AI, but those in hospitality, sanitation and cleaning are least expected to be adversely affected by automation.
Conclusion: Australian business owners and operators – whether at the small end of the SME sector or those responsible for the function of corporations with annual turnovers of more than $100m -- should be considering as a matter of urgency how gen AI can enhance their bottom lines.
However, this should not be done without an understanding of the technology’s limitations, without willingness to involve staff in use of the technology and resolve to consistently upgrade employees’ skills in developing AI content in the context of their business’s needs, current and future.
