The Tax Prep Career Bet Nobody Is Making

The numbers don't add up unless you're on the inside.

Denver's metro has 1.67 million workers. It graduates about 600 accounting students a year. The average tax preparer is over 55. At least 2,000 retire annually, probably more. The pipeline replacing them runs at roughly a third of the outflow rate.

If you heard those numbers about any other profession, you'd call it an opportunity. But mention tax prep and people say "AI." Like the robots are about to file everyone's Schedule C while we sleep.

Here is the thing. AI does the opposite of what people assume. It doesn't shrink the market for tax preparers. It floods it.

AI is a demand engine, not a replacement

Every office administrator AI displaces becomes a virtual assistant with 1099 income. Every junior accountant becomes a fractional bookkeeper with quarterly estimates and a home office deduction. Every paralegal becomes a contract reviewer getting 1099s from three platforms. The simple W-2 is dying. The chaotic multi-source return is becoming the default.

In Denver, somewhere between 80,000 and 150,000 workers are going to get pushed out of entry-level cognitive roles into gig work and self-employment over the next five years. Every single one of them becomes a filer who needs more help than they did before. More forms. More schedules. More judgment calls about what's deductible and what's not.

AI can generate the forms. It cannot sign them. It cannot sit across from a client who's panicking about an audit letter. It cannot carry liability when the IRS comes asking. These are not minor caveats. They are the entire job.

The credential is cheaper than you think

The barrier to entry is weirdly low for something with this much runway. The Enrolled Agent credential, the highest one the IRS issues, does not require a college degree. It does not require an accounting background.

You need a PTIN. Fee: $18.75. You apply online.

Then three exams: Individuals, Businesses, and Representation/Practices. You have three years to pass all three. A background check and Form 23 after that. The prep materials run a few hundred dollars, not the thousands that CPA review courses charge. You study at your own pace.

If you want to start faster, the AFSP (Annual Filing Season Program) is the lower bar. Eighteen hours of continuing education, a six-hour federal tax refresher with a test, and you are listed in the IRS directory with limited representation rights. You can finish it before the next tax season.

Compare this to a coding bootcamp where 6 out of 21 grads in one tracked cohort got jobs. Or a master's degree that costs $40,000 and drops you into the same AI-exposed market you're trying to leave. The tax path has a defined finish line, a government credential, and a supply-demand curve that bends your way.

The guild moat

There is a concept I keep coming back to: the guild return. When university degrees stop signaling competence, what replaces them is not a shinier degree. It's a model where a human with standing validates another human directly. The senior preparer signs off on the junior. The liability attaches to a person, not an institution. The gate is "can you do the work," not "do you have the paper."

Tax preparation is one of the few professions where this is happening now, at scale. The IRS is not going to automate itself out of existence. An AI will never sit across from a client and say "you need to document this differently or the auditor is going to dig in." Clients come back because they trust a person, not a dashboard.

These things get stronger as AI improves. The tools get sharper. The moat gets deeper.

The fear is the filter

I almost talked myself out of this path. The fear isn't about difficulty. It's about picking wrong. That AI headlines would turn out to be right and the whole profession would collapse the month after I committed. That seasonality would make the income unstable. That the ramp was too steep.

Every potential entrant has the same fear. And every potential entrant without a mentor, a client base, or a clear succession plan walks away. The doubt itself becomes a barrier to entry. It keeps supply constrained while demand keeps growing.

The person who reads past the AI panic and runs the actual numbers walks into something asymmetric. Fewer preparers every year. More complex filers every year. A credential that compounds instead of decaying.

The practical sequence

Get your PTIN first. $18.75, online, takes an afternoon. It's not a credential, but it makes the thing real.

Do the AFSP next. Eighteen hours of CE from an IRS-approved provider, including the six-hour refresher with a test. This gets you into the public directory with limited rep rights. You can finish this before next tax season.

Work a season somewhere. H&R Block hires seasonal preparers with a PTIN and basic training. One season teaches you simple returns, and more importantly, whether you actually like the work. That's worth more than any amount of research.

Then decide on the EA. The three exams are real. But they are self-studiable and they unlock unlimited practice rights before the IRS. The credential doesn't expire as long as you maintain it with continuing education. The demand side is not going to stop growing.

You don't need to be the smartest person in the room. You just need to notice that everyone else is looking at the same AI headlines and drawing the wrong conclusion.

If you want help studying for the EA exams, check out the EA Coach — it's a free tool I built that grades your practice answers the way the real exam does.