There’s been considerable talk lately, about AI and its potential to transform HR and payroll.
But how is it actually being applied?
A recent study by YouGov and Deel found that while the majority (61%) of HR decision makers believe AI will positively impact HR over the next five years, less than two-in-five (38%) currently employ AI in their processes and workflows.
Where AI truly has the power to excel is in the handling large datasets and making sense of them.
In the payroll world, there is extensive legislative and compliance data to take into account.
And, laws and requirements change constantly.
AI models should be used to help companies stay up to date on local country rules and policies, dynamically adapting to remain current. And yet the fact these technologies are not being adopted is disappointing.
What we really need to do close the knowledge gap and move beyond AI for merely responding to HR queries (ie through chatbots), and onto more complicated, but core areas.
For instance, it’s far more powering when AI is used to address bigger payroll concerns, like addressing country compliance issues or spotting payroll errors before they become real problems.
Most payroll systems aren’t yet leveraging AI in the right way, but they can and should be.
In fact, I would argue the best uses of AI for payroll in particular have yet to be appreciated or even realized.
Here are the key areas where I believe AI has the real potential to transform a company’s payroll and HR system, while also improving the workforce experience:
One common challenge for employers is explaining to their teams the variations in their paychecks cycle-to-cycle.
G2N variance, as it’s known, is a fairly common monthly occurrence.
It could be caused by expense reimbursements, bonuses or commissions, a change in employee benefits, or a change in the foreign exchange rate.
More extensive G2N variances can occur at the country level.
These might be due to specific government mandates or policies, such as 13th-Month Pay in the Philippines, or the new 2024 tax legislation in the Czech Republic that applied to non-monetary benefits.
It has traditionally been the job of a payroll manager to analyze these variances.
But this task can now be handled by AI.
Trained AI models can determine and add reasons for a paycheck change in a certain country, no matter what the root cause is.
The goal is to proactively deliver to employers and their teams the line items of each percentage variance, ensuring an accurate and precise payroll report and eliminating potential confusion.
Expenses are a headache for payroll managers and workers alike, especially in a global company where expenses are treated differently based on rules and laws within each country.
This is where AI can help in a big way; by removing the potential for error and compliance risk.
For instance, AI can review receipts and invoices against each local market requirement to ensure compliance.
It can also compare them against all historical submissions and a third-party database of documents to prevent accidental duplication and fraud.
On the worker side, AI could alert someone when they’re submitting an expense if it’s above the per-diem limit in a specific country, or if it’s not permitted under local tax rules.
It could even tell a team member why a receipt was incorrectly submitted and show them an example of how to correctly submit it.
When AI is fully integrated into platforms, it can successfully unburden HR and payroll managers by automating and streamlining processes.
Much of payroll has to do with manual customization, and the more AI can automate those customizations and information flows, the better.
For instance, in the backend of PaySpace’s native calculation engine, we’re currently testing AI OCR (optical character recognition).
It interprets PDF documents provided by the revenue authorities, delivering all the report templates required by that specific authority.
Our engineers can then map the relevant fields in the system’s back-end to the required fields on the template.
Eventually, developers will no longer need to map and configure templates, reducing the resource requirement to build localized reports.
Of course, AI chatbots are useful for helping employers get quicker answers to complex, country-specific questions like “What is the parental leave policy in the Netherlands?” or “How many contractors did I hire in Europe this year?”
Eventually, they’ll get so good that they’re essentially “everywhere AI” in the product.
Here, instead of navigating to the chatbot destination, a payroll manager will simply be able to ask a question relevant to the task in front of them.
Such capabilities could also benefit employees, who could ask questions about a deduction on their paycheck to understand the reason for it, or inquire about a certain benefit package and its details.
Why does all of this matter?
Payroll is a company’s biggest expense month-to-month.
Improving processes, giving teams faster access to important information, ensuring country compliance anywhere in the world – these are business-critical efforts.
AI has the potential to be a company’s key partner in making payroll and HR work better — but only if it’s tailored to the real issues facing payroll managers.
