How AI Will Transform Accounting: A $100B Opportunity Explained

Seventy-five percent of accountants are set to retire in the next 15 years, according to the American Institute of Certified Public Accountants. Who will fill these roles and how? In this discussion, a16z investors Seema Amble and Marc Andrusko explore the dramatic transformation of the accounting industry with host Michelle Kung. Despite being a $100+ billion market with over 3 million accounting professionals, the industry still runs on Reagan-era software. Today, the accounting industry remains slow to adopt AI, alluding to challenges and opportunities. This conversation explores why accounting is reaching an inflection point amidst labor shortages, increasing workloads, and new AI capabilities — and what this means for startup founders.

Summary:
00:00 Market context: the vanishing accountant
01:42 Three million professionals & $100B+ opportunity market opportunity
03:26 Why accounting is different: text vs. numbers in AI
05:58 Implementation challenges: accuracy demands and billing models
08:51 Applying LLM strengths to accounting
10:44 Why you should sell through accounting firms
12:00 Promise for consumers to get better and cheaper services

[jaredspencer]

All respect to these guys, but the reason Accounting AI might be slow is that they have a fundamental misunderstanding of accounting. "It's all about numbers, unlike law." Very false; the majority of accounting work is classifying transactions or assets based on unclear government guidance paired against unique business operations. "The numbers have to be 100% correct 100% of the time." Also false; they need to be correct within the limit of "materiality", which might actually work in favor of LLM's; if they hallucinate every ten-thousandth transaction, that would still be better than human systems, and within materiality limits.

[tpg]

I also disagree with the claim that accounting is about numbers. It is about exercising professional judgements to determine how to classify those numbers. Usually it is about the classification of the transactions under both accounting standards and tax law.

[TheDoubleMvp]

Timely video. Basis AI just announced a Series A today, building AI agents for accountants

[tpg]

I wouldn't agree that Reagan Era software is still being used. I've being doing accounting since 1988. Back then spreadsheets didn't even have tabs. You couldn't just download your transactions from the bank and have it automatically reconcile. If there were ETL systems back then they weren't on every user's desktop like Power Query.
LLM's are still of great value. I can translate a German Tax return to English, I use it to record transactions from paper bank statements. It was pretty hard to do with OCR software. I can easily generate code if I need to. I can cook up new functions and behaviors in Excel.

[Anurag-csjs]

Great vid. Have a feedback though, , , , the way captions show up is way too distracting ! For a non native speaker, I end up reading them a lot, even though the content is very clearly put across

[travelinggreek]

AI will definitely ease accountants' pressure by handling repetitive tasks that take up a lot of time.

[Data_CPA]

I love when tech people discuss accounting. I'm a CPA that has moved into coding around AI. I can say 90% of accounting can be done generative AI agents. Unfortunately, the most pivotal aspects are human decision making. The real opportunity is building your own accounting infrastructure for clients and integrate blockchain, AI and data science. Accountants are literally training the AI for software companies and paying subscription fees that lock you into the platform.

[larryrichards]

Reagen era software bla bla bla.. Did you ever work on consolidation, mergers and know about how each firm and task is different on these matters and require human capital with accounting knowledge. Yes there is certain section of financial reporting and audit is repetitive and being taken care of automated “robotic” process. However, when you promote this for the entire profession. It is wrong. You are either not that knowledgeable about this profession or want to promote your product so much. Accounting is using Reagen era software huh!! The most irrelevant video on accounting profession.

[reinhardtbezuidenhout]

Another issue is also the fact that accounting firm's clients use multiple different accounting software and use that in very different and personalised ways. Resulting in the source data to be very difficult to work through and understand, and most of the time there's no real pattern to it. So the data that is fed to AI for processing is actually inherently difficult to train ai on...

When we see AI in accounting, it will be in the reviewer seat - where data has alteady been understood and put into a sanitised format. As opposed to the seat of the person actually doing the work. This is inline with the pattern we've seen AI utility follow - it replaces work from the end of the process first.

[moazzamjadoon]

I totally agree with people who say accounting is about classifying transactions. Machine learning would be of great help However I don't know any ML models that can 100% accurately classify things after they have been trained. A 90% accuracy would even be a challenge. Eve deep learning models may not be up to the challenge. However, AI can greatly speed up the drudgery of more mundane and routine transactions. You will need a person to constantly audit the entries by AI. Also, the Business Process Management models can be used to help AI with routine entries and leave less routine entries, such as Long Term Assets classification to humans. What I mean is, for example with BPM model in place, the routine sales entry can easily be left to the AI. We can use Process Mining to check if these routines and routine entries are being accurately entered.
Lastly, I would like to say
While AI isn't perfect, its ability to handle repetitive tasks at scale can free up human resources for higher-value work.
Also, BPM and process mining not only help ensure accuracy but also offer a feedback loop for continuously improving AI models.

[kornflakesjunkie]

awesome. but please remove these distracting subtitles in the video.
the ones we can activate are sufficient.

[rkusri]

Hmm...didn't realise a16z also has a 'moderator' panel to delete comments that counter this hubris

[clementgavi]

Financial statements are obtained via transactions recorded by bookkeeping, the basic accounting operations,
Let's imagine a code bar on a product sold in a shop for 100. The information of this transaction refers to inventory, account of class 3, customer, account of class 4, cash, account of class 5, sale, account of class7. etc. The balance sheet is concerned with three accounts here, 3, 4 and 5, and Income statement with account 7.
If AI can make a code bar that include all these information and generate financial statement accordingly, the salesperson who serve the customer does the job. Because there is no more need an accountant to record the transaction, etc. The same goes with the analysis of the statements.
Anyone who understand how the statements come about, understand that technology can transform accounting.

[dertynan]

Why is A16Z so out of touch, these days? Accountancy is first and foremost about accuracy. AI hallucinations and errors are well-documented. How many CFOs want to sit in front of a Senate Subcommittee because the AI doing their accounts made a grievous error?

[Jonathanx]

Its amazing how software testing which is soo repetitive is not automated. Google, Microsoft anyone?

[kjjjji]

This people are behind the technology trends. AI has been solving Accounting problems for over a year now. It was actually the first test l tried as soon as Open AI started releasing their models. Been using it for taxes for two years now, even for financial services advising. Keep talking yourself out of your job.

[mrbabyears]

This sounds good in theory. Yes, Ai will make accounting more efficient. Accountants will just do more with less. The only the price wars happen in accounting is if there were an over abundance of accounts that used Ai. Then they have have to compete for business. But this isn't the case. This is interesting. You would think, technology would make things cheaper. There are so many things that should be cheaper due to technology advances but this isn't the case. Infact things have become more expensive because companies pass those profits to their top executives and their share holders. Companies use technology to enrich themselves, cut jobs, eliminate pensions, etc.

[theshaneburns]

IRS expense deduction categorization is the most abundant, thoughtless work in accounting.

[embarq]

Automation is the need. Not necessarily using Gen AI but even traditional AI / machine learning can be useful

[amishthaker]

Feedback - the subtitles are SUPER annoying. If I want subtitle I will turn them on on youtube. Why reinvent the wheel?