The Ticking Time Bomb of AI in the Professional Services Industry

Overview

According to the Bureau of Labor Statistics, the professional services sector represents around 13% of the US Gross Domestic Product (GDP). This sector, oftentimes referred to as “knowledge workers,” includes companies such as law firms, marketing agencies, business consulting firms, accountancies, IT services, and software developers like CodeStringers. Up until a handful of years ago, no one would have ever thought that these knowledge workers would be at risk of losing their jobs to machines. But that day may not be as far away as you think.

How Do Professional Services Companies Work?

The typical way that professional service firms make money is that they have a staff of professionals (lawyers, accountants, etc.) who perform services for the company's clients, and they then bill those clients for the worker’s time. Usually they charge by the hour, but sometimes it's bill-by-project, or sometimes it's a combination of the two. They try to bill their clients 3x the amount that they pay those employees so that they can cover their fixed costs and their customer acquisition costs and (hopefully) still have some profit left over to compensate them for the risk involved. So the equation looks something like this:

Profit = (Billable Rate / Hour - Labor Cost/Hour) x # of Hours Billed - Customer Acquisition Costs - Fixed Costs

Over time, the profit number has to be > $0, otherwise the company goes out of business.

What Has Changed With AI?

There are a handful of dates in history that I have lived through which I consider to be “days that the world changed.” One was 9/11, one was the day the Rodney King verdict was read, and one was 30 Nov 2022 when ChatGPT came out. Why was that date so significant? Because it was the date in which the human race proved that they could develop an application that could do the job of a knowledge worker. And it only gets more powerful.

As of the date of this writing, AI can pass the bar exam. In fact, OpenAI’s GPT-4 has reportedly scored in the 90th percentile on the Uniform Bar Exam, outperforming many human test-takers. The Turing Test has been passed with flying colors. AI can now do the job of almost any professional service worker. See below for just a handful of examples of what it can do (limiting the examples to professional services only, the list would go on for pages):

- Design software applications

- Write and QA software code

- Write legal contracts (e.g., Allen & Overy’s “Harvey” AI tool)

- Design/draw logos

- Create print and digital ads—both art and copy (e.g., Ogilvy’s use of Midjourney and ChatGPT)

- Write scripts

- Animate video clips

- Audit financial statements (e.g., PwC & Deloitte’s use of AI in auditing)

- Translate from one language to another (e.g., Unbabel AI-powered translation)

This is quite an extensive list, but is AI good enough yet to completely replace a knowledge worker? The answer is no. AI still makes mistakes, and it tends to recycle old ideas instead of inventing new ones. However, one thing is for sure: it definitely can cut down on the number of workers you need to do a certain amount of work. For example, a legal contract that used to take 100 lawyer-hours may now only take 10. Case in point: DoNotPay, the so-called ‘AI lawyer,’ has been used to draft appeal letters and simple contracts in a fraction of the time it would take a junior associate.

Why Is This A Problem?

Remember the profit equation for professional services companies? The math gets scary when you take into account what AI can do. Let’s take the example of the law firm of Dewey Cheatham and Howe, LLP. And let’s assume, conservatively, that AI can cut down on the number of billable hours per project by half. Here’s what happens to their business model:

Revenue per Client

See the problem? Now instead of making $18k per client, they’re now making $3k per client, and this is BEFORE fixed costs are paid. Needless to say, this is a major problem.

We’re already starting to see the results of this margin-compression pan out. For example, the 2025 SPI Professional Services Benchmark found that revenue growth in the sector slowed to 4.6% and EBITDA fell by 36%, highlighting mounting financial pressure. According to Business Insider, AI-driven automation could eliminate up to 50% of jobs in audit, tax, and advisory roles across the Big Four firms within the next few years.

So What Should Professional Firms Do About It?

Mathematically, there are only 6 options that are possible if a professional services firm wants to survive this margin-crunch:

1. Increase their hourly rate: Prices are set by supply and demand so, theoretically, one firm does not have the ability to increase their prices by much without risking losing their business to a competitor. So we can assume that this will not work.

2. Increase their number of clients: This is an option available only to large firms because smaller firms aren’t going to have the budget necessary to increase their marketing budget enough to attract this new business. So this option will only work for some.

3. Decrease labor costs: Current employees aren’t going to want to take a big pay cut. Therefore, drastic measures would have to be taken, such as offshoring or hiring more early-career workers. So this option may work a little but will be painful.

4. Reduce fixed costs: This will also be tough in the short term, but it will be a necessary step if you can’t find any other options. So downsize office space, reduce travel, reduce non-essential non-billable employees. This option is possible and necessary.

5. Reduce acquisition costs: You’re welcome to try, but this is unlikely to happen, especially if you’re trying to increase your client base. So we can assume that this will not work.

6. Find other sources of revenue: Maybe there is a way you can find revenue streams that aren’t susceptible to this sea-change? For example, McKinsey & Company has launched AI learning platforms and masterclasses to monetize its internal expertise, and Clifford Chance offers AI-focused training programs for clients. This option is difficult but possible.

If you can’t make one of these above options work, then your business will not survive.

Conclusion

I know this is scary. To be honest, it kept me up for a few nights straight the first time I realized that I was one of the ones who was going to be affected by this change. However, I eventually shook off the depression and realized that I needed to rise to the challenge.

In the early part of the 19th century, there were a group of angry textile workers who were angry about the introduction of machines into the textile-making process because they knew that it would reduce the number of textile workers needed. So, in response, they decided to destroy the machines so that the firms would hire them back. Did it work? No. They not only lost their jobs but the ones that were caught were either hanged or sent to Australia’s harsh penal colonies. They were called the “Luddites.”

Today, the term refers to anyone who tries to stop the advancement of technology because they’re scared it will affect their livelihood.

So don’t be a Luddite.