Somewhere around 2023, every finance headline started sounding like a countdown clock: robots were coming for accounts payable, and accountants supposedly had maybe five good years left.
It’s 2026. Check your Accounts Payable department. It’s still there, still human, and still catching the invoice that got coded to the wrong cost center.
So what happened to the AI takeover in accounting?
Where Did the “AI Will Replace Accountants” Idea Come From?
The fear wasn’t irrational, it just got the wrong job title. AI genuinely is automating a specific slice of finance work: high-volume, rules-based, repetitive tasks.
The U.S. Bureau of Labor Statistics is direct about this: employment for bookkeeping, accounting, and auditing clerks is projected to decline 6% from 2025 to 2035, largely because software has automated many of the routine tasks these workers perform.
That’s a trend describing the automation of routine clerical work, not the disappearance of accounting judgment.
And AP/AR sits right at the seam between the two.
How AI Improves Accounts Payable and Accounts Receivable
Let’s give credit where it’s due. Modern AI and automation tools are genuinely good at:
- Matching invoices to purchase orders
- Pulling data off scanned documents (OCR)
- Flagging duplicate or anomalous transactions
- Handling routine reconciliation
This is the grunt work layer, and offloading more of it to AI is why it’s freeing up AP/AR staff for higher-value work, not eliminating them altogether.
What’s Still Stubbornly Human
On the AR and collections side, actually collecting the money still has to be a person’s job, says Lindsay Pfenning, Stivers’ Chief Operating Officer.
“Employers still need humans to make outreach calls, review inconsistencies or potential errors, and research invoice or payment disputes until they’re resolved.”
Lindsay Pfenning, Chief Operating Officer
A slow-paying client, for example, could be dealing with a cash-flow problem, a billing issue, or dissatisfaction with the service. AI can flag the overdue payment and even help with the initial outreach.
But figuring out why the payment is late and how to handle the situation still takes human judgment and empathy.
Accounts Payable has its own version of the same challenge. The software can handle standard payable processing just fine. But when something doesn’t match, someone has to figure out why:
- an incorrect invoice amount
- a deduction that doesn’t match the PO
- mismatched documentation
- a vendor dispute that requires an actual conversation
Augmented vs Automated Accounting: What’s Really Happening
The employment data backs this up more than the headlines do. BLS projects accountant and auditor employment will grow 5% from 2025 to 2035 — faster than the average for all occupations — with roughly 115,300 openings projected annually.
Pfenning’s read from the field lines up with that number: competition for accounting talent overall is fierce right now, and that squeeze extends to bookkeepers and clerks who carry AP/AR responsibilities, too.
Part of that demand shift comes down to what the job actually requires now. Traditional AP and AR roles used to be, in large part, data entry. AI and system automation now handle a lot of that.
Pfenning points out that some longtime AP/AR staff nearing the end of their careers are taking early retirement or seeing their roles eliminated as the work changes beyond what they’ve historically been doing (a scenario Stivers has helped clients navigate directly when a finance leader’s early exit created an unplanned gap).
What’s replacing it demands a higher-level skill set: critical thinking, the ability to manage a higher volume of work with the tech doing the heavy lifting, and sharp cash management — especially in the current economic climate.
Read that next to the BLS clerk-role automation trend and the picture gets clearer: it’s not “AI vs. accountants.” It’s routine task volume shifting to software while judgment-heavy work — including a lot of what AP/AR specialists actually do day to day — still needs a human. Just a differently-skilled one than a decade ago.
What This Means If You’re Hiring for AP/AR Right Now
If you’re building out or backfilling an AP/AR team, the calculus has shifted, not disappeared:
- You still need people. The judgment-call work isn’t going anywhere.
- You need different people. Comfort with automated tools and exception-handling instincts matter more than manual data-entry speed now.
- The accounting talent shortage makes this harder to solve alone, especially for roles where “figures it out fast” is the unwritten part of the job description.
That’s usually where a staffing partner who knows the accounting and finance market earns their keep. Not by filling a seat, but finding someone who can actually do the judgment-call part of the job the software can’t.
Where AP/AR Roles Are Headed (Not Where the Headlines Say)
It’s not a shrinking function. It’s a shifting one.
The people who’ll thrive in AP/AR over the next few years are the ones comfortable working alongside AI tools, not competing with them: reading the exception reports, owning the vendor relationships, making the calls the software flags but can’t resolve.
So yes, the robots showed up.
They just took the boring parts of the job, like everyone secretly hoped they would.
Building out an AP/AR team that can handle what the software can’t? Let’s talk.
