By Katherine Boychick, Chief Technology and Innovation Officer of the EY Region
In the Australian boardroom, executives ask, “Is it safe?” Every time an AI implementation occurs. The rest of the world competes with artificial intelligence, but Australians tap on the brakes.
The recently released EY Global AI Sentiment Index shows that the fear of cybersecurity and the lack of trust are two main reasons why we are behind the rest of the world. Our financial future outcomes can be serious if we don’t address this head.
There is a breakdown of trust. Australians are deeply skeptical of AI security, with 74% ranking security obstacles as their biggest concern, far surpassing the 63% global figure. This security anxiety lowered the overall AI sentiment score to 54 out of 100, but the global average is at a much healthier 70.
Even more impressive, 80% of us are worried about fakes and false reports generated by AI. We are more concerned about this than most other countries we are surveyed. This explains only 37% of Australians who believe the benefits of using AI outweigh the risks compared to the global 51%. This shows a breakdown of trust focused on specific security.
The fear of being deceived runs deep. Content generated by sophisticated AI is difficult to distinguish from human-generated material, so Australians wonder what this means for information integrity.
For businesses, this is even more concerning. Imagine an AI portrayed by executives who allow fraudulent transfers, manipulate video calls that give false instructions, and create operations that create communications that redirect payments. These are not overstated scenarios, they are real threats that awaken cybersecurity experts at night.
Unlike security threats that target target systems, AI-powered attacks target the target trust itself and rebuilding that trust is much more difficult than restoring compromised data.
Our study also suggests a generational disparity in which 63% of Australian Generation Z and 60% of millennials say they are used to AI, as opposed to 48% of baby boomers and 52% of Gen X.
I put this into hard-earned wisdom. Those who have witnessed decades of change often ask the most insightful questions about security. This question deserves a proper answer, not a fire.
These concerns need to be implemented in strict governance and security strategies to address them.
However, there are economical prices to hesitate. If only 16% of Australians fully understand AI (a small improvement from 13% last October), there is a knowledge gap that exacerbates the fear of security. Less than half of us (48%) feel comfortable using AI every day.
Don’t make a mistake – this gap threatens Australia’s economy. While we are careful and deliberate, our global competitors are moving forward with AI and our position as laguards can have a hard time keeping up. Security concerns that hinder adoption today will be more vulnerable tomorrow as they are ironically rushing to implement unfamiliar systems under competitive pressure.
There are five practical security approaches that work.
It shows you exactly how to protect your data, limit access, and store integrity. Document this in plain language that non-experts can grasp.
Regular exercises and third-party security audits are not just good practice, they are powerful trust builders.
Misinformation is a major concern for Australians, and therefore, investments should be made in both detection techniques and verification protocols, especially in high-stakes situations.
Security education significantly increases the trust of AI across generations. It focuses on practical, role-specific training that people use every day.
Develop and share communication plans in security crisis. Nothing builds trust that shows you’re ready when things go wrong.
The growing security awareness in Australia doesn’t have to be Achilles’ heels. That could be our strategic advantage – if we lead it productively. By developing security-first AI systems, Australian organizations can build powerful and reliable technologies.
The link between security reliability and AI ingestion is unmistakable in data. Organizations that have successfully closed this gap are not what they treat cybersecurity as a compliance checkbox, but as the basis of their strategy, especially when it comes to AI.
The views expressed in this article are those of the author, not Ernst & Young. This article provides general information and does not constitute advice, and should not be relied on in that way. Professional advice should be requested before relying on any of the information. Liability is limited by schemes approved under professional standards laws.
