For all the noise around artificial intelligence, it still isn't making much of a dent in payroll — at least not yet. A new Zoho Payroll survey of more than 100 US payroll professionals found that only 7% of payroll teams consider AI central to how they work, while nearly half — 44% — say they haven't started using it at all.

When teams that are holding back were asked why, the most common answer wasn't cost or security. It was simpler: 30% said they don't yet understand how AI can actually help payroll. That knowledge gap, more than skepticism, is what's keeping adoption low.

Adoption Is Still in the Early Innings

The picture the survey paints is one of cautious curiosity rather than transformation. A small minority have woven AI into their core process, a large share haven't begun, and the biggest barrier is education, not resistance. For vendors and HR leaders alike, that suggests the near-term win is showing payroll teams concrete, practical use cases — not selling them on the technology in the abstract.

What Payroll Actually Wants AI to Do

Asked which AI capabilities would have the biggest impact today, payroll professionals were clear about their wish list. Catching payroll errors before a pay run processes topped it decisively at 66%. Next came delivering real-time compliance updates (36%), answering employees' payroll questions (31%), intelligently flagging compliance risks (20%), and auto-syncing journal entries to accounting (15%).

Tellingly, predictive analytics — a capability vendors love to spotlight — finished dead last at just 8%. The message is that payroll teams are far less interested in forecasting future costs than in preventing today's problems and cutting the manual work in front of them.

---

For HR and payroll leaders, the takeaway is grounded and actionable: the most valued AI isn't the flashiest. It's the quiet kind that stops a mistake before it reaches an employee's paycheck and keeps compliance current without adding to the workload.