One in three Australian workers is using AI tools at work without approval, according to a new Employment Hero study of 2,600 employees and employers.
The numbers: 75% of workers say AI improved their productivity. 74% report better work quality. But 33% are using it in secret, and 44% of employers say staff are running company data through personal AI accounts.
The guilt factor is real: two in five workers say using AI feels like cheating. More than half are self-training with online materials because their companies have not provided guidance.
What this means for sales teams
If your reps are using ChatGPT to draft outreach emails or research prospects, they are probably doing it on personal accounts. That means your pipeline data, customer intel, and competitive research could be sitting in ChatGPT's training data unless they are on an enterprise plan with data protections.
The productivity gains are legit: AI tools can speed up prospecting, email personalisation, and CRM admin. But without clear policy and proper tooling, you are creating compliance risk and missing the chance to standardise what actually works.
The ANZ context
The National AI Centre reports 49% of Australian organisations are adopting AI, but 19% say they are still unsure how to use it. That gap shows up in this study: workers are ahead of policy.
For sales leaders, the move is straightforward: give your team approved tools with data security, train them on what works, and stop making them feel like criminals for using technology that demonstrably improves their output.
The alternative is what is happening now: one third of your team is using AI anyway, just without oversight, training, or data protection. That is not a policy, that is a liability.
Worth noting: Employment Hero did not break out results by role, so we do not know if sales teams are adopting faster than other functions. Anecdotally, sales professionals tend to be early adopters of productivity tools, so the 33% figure is likely conservative for GTM teams.