Will AI Actually Take Your Job? What Automation Means for AP, Credit Control and Bookkeeping Roles
"Will AI take my job?" It's come up in more candidate conversations this year than salary has. Ask it about accounts payable, credit control or bookkeeping specifically, and the honest answer isn't yes or no — it's that the job you're doing in three years won't look like the job you're doing now, and the difference matters a lot more than the headline.
So let's skip the vague reassurance and the doom-scrolling, and actually look at what's changing, what isn't, and what to do about it if you're the one doing the job.
What's actually changing in AP, credit control and bookkeeping
Accounts payable is the furthest along. According to The Hackett Group's 2026 research, AP is now the single most mature finance function for AI adoption — 33% of organisations are already scaling AI solutions in AP specifically, well ahead of the rest of finance.
And it's not just data extraction anymore. Forrester's 2026 analysis of AP invoice automation points to a shift from AI that just reads and codes invoices toward "agentic" AI that handles exception management, fraud detection and supplier queries with far less human sign-off than a year ago.
Credit control and bookkeeping are moving too, just less dramatically — automated matching, chasing and reconciliation are further along than most people realise, even if the headline-grabbing "agentic" tools are concentrated in AP for now.
The honest numbers on how people actually feel about it
You're not imagining the anxiety, and you're not alone in it. ACCA's Global Talent Trends 2026 report — the accountancy body's own annual survey — found 51% of finance professionals are worried about AI's impact on their own job, and it's younger, more junior people who are most concerned. That tracks: junior, task-heavy roles are exactly where AI has landed first.
UK-wide, CIPD data shows a similar pattern from the employer side. Among UK employers who expect AI to cut headcount over the next year, 62% named clerical and administrative roles as the ones most likely to go — by far the largest share of any category. Overall, 17% of UK employers said they intended to reduce headcount because of AI, against just 6% expecting to increase it.
Those numbers are real, and we're not going to soften them into "AI will create new opportunities" without saying what we mean by that.
Where the automation stops and you start
Here's the part that actually matters day to day: AI is good at the parts of AP, credit control and bookkeeping that are repetitive and rule-based — matching, coding, chasing, first-pass reconciliation. It's much weaker at the parts that involve judgement calls, awkward conversations and knowing when a rule shouldn't apply.
A credit controller's actual value was never "can type into a system and send a reminder email." It's judging which client relationship can take a firm phone call this week and which one needs a different approach, and knowing that from the account history, not a script. An AP clerk's value shifts from keying invoices to owning the exceptions the system can't resolve — the mismatched PO, the supplier dispute, the one invoice that looks fine but isn't. A bookkeeper's value moves from data entry toward reviewing what the software produced and catching what it got wrong.
None of that disappears when the repetitive layer gets automated. If anything, it becomes a bigger share of what the role actually is.
What this means if you're in one of these roles right now
If you're in accounts payable, the practical move is to get comfortable being the person who handles what the system flags, not the person who processes everything by hand — employers are already hiring for that shift, not against it.
If you're in credit control, the relationship and negotiation side of the role only gets more valuable as the admin side automates — lean into that rather than treating it as a separate skill from "the real job."
If you're in bookkeeping, the direction of travel is toward review, oversight and client-facing explanation of the numbers, and away from pure data entry — worth reflecting in how you talk about your own experience at interview, not just in the job you're doing.
What's actually worth doing about it
The most useful stat in ACCA's report isn't the 51% who are worried — it's that 82% of finance professionals say they're confident they can learn and apply AI-related skills when they need to. That's not blind optimism, it's people who've already adapted to a decade of new finance systems doing the maths on the next one.
Concretely: get hands-on with whatever automation tools your current employer already uses, rather than waiting for a course to tell you it's safe to start. Ask to be the one who handles exceptions, not just volume. And when you're job hunting, ask a prospective employer directly what they've automated and what they haven't — the answer tells you more about the actual shape of the role than the job title does.
We place people into AP, credit control and bookkeeping roles across Coventry, Birmingham and Warwickshire every week, and this is the question we get asked most often right now. If you want a straight answer about how a specific role or employer is actually using automation — not a guess — that's exactly the kind of thing we know because we're in these conversations constantly. No waffle, just what we're actually seeing.