AI Taking Accounting Jobs
Just in time for some students. :)
Hello from me!
I have a chat tonight with some lovely students from my alma-mater and the professor there, Dr. K., had the students submit some questions. Those quesitons can be boiled down to 6 themes
AI Taking Accounting Jobs
Crypto / Blockchain / Web3 = Curiosity + Confusion
New Career Paths: They Want Permission to Not Be Traditional
Skills & Certificates: “What Do We Learn NOW?”
Macro: “Where Is Accounting Going?”
Regulation, Fraud, Controls, Governance
This memo is to help address some concerns around these questions, though I admit I think I can only answer that partially.
Cheers y’all!
p.s. I have a lot to share, I hope to get to that soon.
AI Taking Accounting Jobs
The #1 set of questions being asked by students this time around is focused on AI & Accounting. Some can see it as taking jobs, where as if put my hat on and pretend I’m a student, I see it as a source of anxiety. Rightfully so!
There were over 50+ questions asking at least one of these:
“Will AI replace accountants?”
“What tasks go away?”
“How will entry-level work change?”
“Will job opportunities shrink?”
“What skills keep us competitive?”
“What human abilities still matter?”
“How should students position themselves?”
And the anxiety students (me, I don’t know about the students) feel isn’t entirely irrational. It’s very rational, born out of a generation of students who are looking at a career path that was invented in like 1980, kinda digitized somewhere in the 2000s, somewhere in there experienced outsourcing, and now has to consider AI in the 2020s.
What is important here is “why” are we having these thoughts, and often time, it’s due to a competitive job market, especially for new graduates, and the pursuit of a career that justifies the degree. What does a student’s future career look like? More importantly, what is student’s future look like in terms of earning a livable wage? Is it a stable future, or unstable future? (Three body problem reference)
Contextually to my alma-mater and the socioeconomic dynamics, the answer you probably have heard to your thoughts are unsatisfying, and probably around eliminating jobs and shifting dynamics.
So let me spell out to you the current state of an accounting career, or “old accounting career”.
I expect the accounting career, the old public accounting career, that has probably been conveyed to this demographic via firm speaking engagement, to be eliminated, or made irrelevant. It will not go slowly, but it will go surely. I expect a variation on what we see as the accounting career that focuses its efforts on “why”, “where”, and “when”, and less about “how” and “What”, albeit with 10x the work, with tools that allow you to do it at 10x the speed.
That Old Public Accounting Career
It was going when I was in school, it should have died in the pandemic, and now with AI, it is effectively sealed. At smaller, more historic firms, you will see the by-product of those eras, a time before digitalization.
Perhaps at the Big 4 or the recently acquired Private Equity Company that used to be Moss Adams, you’d see more digitization. But outside of those names, I fully expect this old accounting paradigm to be eliminated, or consolidated.
The old current accounting career is structured like an apprentice model.
These kinds of working dynamics have a “staff” or “associate”, or some other “entry level” worker focused on performing routine, and while they may joke it’s decently true - brainless jobs.
This is code for “I’m going have you do something, and repeat it hundreds of times”, and then to reduce the risk of error, I’ll make sure the senior “redoes your work” so two people did the same task a few hundreds times. Then, to be very sure, especially on higher risk accounts, I’ll have the manager repeat your work too - though they’ll take some shortcuts because I expect the manager to have oddly more niche experience that you get after doing the job a few thousand times.
The Hard Skills of the Old Current Public Accounting
For almost all entry-level public accounting jobs (and yes I cover private too), it comes down to these “hard skills”:
Data Requesting, which is known as PBCs (Provided by client) or Requests Lists, is a age old project management thing that involves an auditor or tax person or accountant asking for a bunch of things, and expecting the client to give a bunch of things.
If you’ve ever tried to prove you are qualified for financial assistance (FAFSA), that you are a US Citizen and can fill out a W9, or tried to take out a mortgage, at some point you were slapped with a “pleased provide X, Y ,and Z”.
Alternatively if you filed your own taxes, and had to get random statements, you know what I mean.
If you haven’t done these things, the best proxy I can offer you is renewing your DMV license and trying to prove you are currently a resident.
If you a foreign student and got a VISA, I guarantee you that was probably harder than most of these things I’m going to list.
Tick & Tie, which is the process of taking a report and basically using the accounting equivalent of citations to a source document.
Bank reconciliations, which is the point-in-time driven procedure of matching bank statements or cash held at third party institutions with the end of the month so we can argue who owns all the cash or not because somehow in 2025, “cash can be in transit” like some weird physical check was cut and sent using the address found in a rolodex.
Sample testing, which is the accounting value add of choosing random things that happened in a “process” or “financial accounting” and trying to prove it happened, or didn’t happen consistently.
Invoice Vouching, a time old fable that involves getting stacks of paper, often a purchase order, the bill of lading (shipping notice), and the invoice, and saying “yep” they match, it happened, no funny business except sorry senior, you have to re-perform what I just did and review this stack of paper. Also sorry I didn’t sticky note what “piece of text” I saw to give me the impression it was good.
Expense Categorization, or basically sorting. You know companies pay people to “sort” what an expense means based on an established chart of account aka we have like 50 choices, make sure the expense is 1 of the 50 choices?
Many auditor busy seasons are based on light or substantive testing that involves burdening their clients with requests for “Artefacts” aka documents (pdfs or paper, depends on industry) so that the auditor can match data-sets together across various systems or processes.
For tax accountants, their busy seasons are tie-outs and make sure their internal docs flow into the 1040 / 1120 tax form appropriately, supporting each financial accounts. If their google-fu / chat-gpt-fu is strong, they will do research to support a taxpayer’s unique case which involves basically going to Bloomberg Tax BNP and finding tax and legal case law to support an argument or credit, or skimming OBBB. Usually these kind of antiques begin because the taxpayer heard from a friend who heard from an instagram reel that there is a new tax credit for XYZ.
The Hard Skills of the Old Current Private/Industry Accounting
And for all the the industry accountants, those not working in public accounting, they have the joy of the following entry level tasks:
Accounts Payable: Pay an invoice based on the contract terms & if the department approves it for payment. Bonus points if they can figure out “if we actually got what we paid for.”
General Ledger: Make sure journal entries for various accounts book into the financial system. Hurt your head when the system bugs out.
Payroll Coordinator: Lock yourself in a room for 2 days and ensure everyone punched in time sheets or that their salaries are flowing correctly based on what HR tells them, and while we’re here, making sure any deductions that the employee made in the HR System flows through. Alternatively, call the HR System (ADP, Workday, etc) and complain their system doesn’t work as expected.
Cost Accountant: Tell us how much something cost based on what was consumed and time spent. Have fun working with legacy systems or various systems. I’ll like you more if I can go to you and you can tell me how much a window pane SHOULD cost.
Billing Analyst: Uh, you bill our customers based on whatever contract they agreed to with us. Send an invoice and the payment link.
Internal Audit Associate: There is a policy or audit plan to “test” each financial accounting or “process” to see if it is “consistent” across a period and you will follow it. Sorry, we no longer give clipboards.
Everything I listed was going to be obsolete the moment we started to digitize the workforce with the internet, and made the grand ole push in the 2010s. In the early 2010s, we talked about Digital Transformation. Digital transformation was going to accelerated the work that could be done, with less resources. The 2020 pandemic accelerated digital adoption at a global level (yay).
AI is a continuation of that energy as companies move to reduce labor cost by reducing the intensity of a problem.
The apprenticeship model used to make sense when the working world was paper driven, systems (computer or processes) didn’t interact each others, and the core essence of accounting was “turning documents and hearsay into numbers with meaning.” Business models were built on that world.
The Next 10 Years
In the next 10 years, that accounting world, specifically the businesses of the old paradigm, will continue getting sunsetted at even faster pace, whether through aging & retirement, through getting acquired, through closure, or through reinvention, and what will remain is what I hope is a new paradigm. One focused on the “meaning” of accounting, as opposed to what does accounting mean.
Well that didn’t make sense but it sounded clever.
What I hope to see is less oddly specific accounting roles, and more thematic basket of roles. Instead of a specialist of 1, it’s most likely a specialist of many - but rather than be an A+ specialist for the topic, I think a passable C grade specialist in a myriad topic if fine. That is the point of AI - to get to an answer that moves us forward, not what gets it perfect.
Here’s what I think the next 10 years will have in store:
Roles will fit the following themes:
I’m here to be blamed.
A lot of professionally licensed firms, whether law or CPA, continue to exists for their ability to certify independently the status of a company.
They exist solely to take the blame. Goldman Sachs is a great example of “if the strategy does great, the company takes the credit, if it it goes bad, you can blame Goldman Sachs.”
A lot of senior roles are like this - they are the “accountability” of accounting, literally.
If you have great soft skills at being relationships, congratulations - this is a great place to be.
I’m here to build operations and make sure we can scale our workforce or technology.
This is the kind of person who thinks in process improvement, technology operations, or scaling.
We also call them a generalist, and if they happen to be great at talking to people and building influence, we’ll call them a politician too.
This one doesn’t really need soft-skills but benefits from being able to make cross functional relationship. But the role itself can be done without such a focus on having soft-skills that make you an excecutive.
I’m not here to lead people, I’m here to be a nerd and work on new implications, or new concepts I don’t know that you deem are technical.
A lot of unusual, one-time transaction types that were complex due to how much energy and resources it took, will become a lot more cheaper and frequent and permeate into different industries. Someone is going to need to understand the tax, or regulatory, or financial implication of that decision.
Alternatively, I’m the tech wizard because my millennial boss who can’t book a flight using a phone and needs to open a laptop goes to me for all the scripting and processes and configs.
This leads to contractor opportunities in the future.
I’m here to understand the business and the context.
This one I like, because of all things I have proposed, it is the most pure.
Accounting is giving meaning to numbers. And what do the numbers mean, is a function of either a very good FP&A professional or analyst, or strategist. Those who can understand a business and all its components, not just from the mechanics, but from the environment, incentives, and challenges, are in for a treat.
Which is hilarious because this brings the profession full-circle again: When someone sees a CPA, they see someone who is good and understands general business. That sentiment has wane the last decade, but perhaps it will make a resurgence.
This could also be the person that is the advocate of value, almost like an appraiser. An appraiser looks at real value, and perceived value - where in AI can help us understand real value, it takes quite a culturally aware human to see the various shapes of perceived value on the buyer or seller of an idea.
Job Titles of the Future
Some I will propose, some I’m going to make fun of.
AI Controller - this is stupid. That’s like saying I’m an Internet Controller in the late 1990s when the dot-com era was around. You are just a controller, and you have various automations at work. Maybe Rillet or Campfire has made processing so fast that the accounting processing portion of the job only takes 10 minutes of a month and you have moved on to different things.
It is possible that the Controller of the future is downgraded from a director/executive to a senior-accountant level role as the expectations of the controller are less.
This also implies you would need to do more.
FinOps Architect - this is good one, because the internet, blockchain, and AI, has bled IT considerations into all other roles. A FinOps architect is then someone who understands systems, both human and digital, and how they do or do not operate together, and thus ideally, would know where the gaps are or where the opportunities could be, and where time is best spent.
Global <insert any title from the past> - truth be told, what was done on a local scale or regional, can be done on a global scale with the same labor with the advent of more technology.
Contract & Config Specialist - Someone who can read both the legal contracts & the associated claims, as well as digital smart-contracts or system logic, to see if what we legally penned with a customer is matching what we have configured, and if there is a deviation - what is the risk and response? Ideally, this role focuses on taking chaotic data managed by other outsiders, into decision making.
OI Analyst - Operational Intelligence or Financial Intelligent Analyst is someone who works on taking data into decision making points. You’d ideally be responsible for the pipeline of data leading, as well as the presentation of data, as well as what is trending.
Payments & Treasury Specialist - This one is fun because “how someone gets paid” and “how to pay someone” will probably still be one of the hardest questions going in to the future due to how people and countries make that complicated.
FinOps Specialist - The irony of this title is that it’s a FinOps generalist, which is effectively someone who can do all the traditional accounting roles mentioned previously, with even less time. Basically, this means do more with less.
A big question I didn’t address is what will entry level jobs look like. I hope to answer this in time, but my hunch today is to take way we know as traditional, and what we think traditional routes are dominating the linear career progression model, and throw it out the door.
I don’t careers are linear, thats a by-product of the Gen X era. I think we believe or have conformed our career stories so that they are linear, but they aren’t.
I shall talk about this another time.