The AI Bubble Is Deflating. Here's Why That's Good for Tax Careers.

2026-07-21

Last reviewed: July 22, 2026. Analysis based on market data as of the July 2026 tech selloff and BLS workforce projections.

The global AI stock selloff that started this week is being called a bubble burst. Chip stocks are leading the decline. Asian markets followed US markets down. "The AI bubble is popping" is the headline everywhere.

If you're studying for the EA exam, or considering it, this is good news.

Not because you should care about Nvidia's stock price. Because the AI hype cycle has been the single biggest source of career anxiety for anyone considering tax preparation. Every headline that said AI would automate accounting, replace preparers, make the credential obsolete. those headlines were powered by the same hype that's now deflating.

When the hype deflates, the fear deflates. What's left is the actual supply-demand gap, which favors preparers regardless of what AI does or doesn't do.

The crash is real

On July 17, the Nasdaq dropped 1.47% led by semiconductor stocks. Asian markets tumbled the next day. Deutsche Bank called it "fresh doubts about the AI trade." The phrase "AI bubble bursting" moved from analyst notes to mainstream headlines in a single trading session.

This follows months of mounting evidence that the AI spending boom was outpacing the revenue it generated. Tech companies have committed $725 billion to AI infrastructure while laying off over 45,000 workers this year, many explicitly citing AI as the reason. AI-driven layoffs are now the leading cause of US job cuts, overtaking restructuring, offshoring, and economic conditions.

The bubble is not just a stock market story. It's a labor market story. Companies are spending billions on AI and cutting the workers AI is supposed to replace. before the replacement actually works. The layoffs are real. The productivity gains are still mostly projected.

What the bubble burst means for tax preparation

The tax preparation profession sits in an unusual position relative to AI hype. It looks vulnerable from the outside. software can fill out forms, AI can classify documents, the IRS has a free file system. But the actual threats are narrow, and the actual protections are structural.

Licensing. Circular 230 governs practice before the IRS. Only credentialed professionals. EAs, CPAs, attorneys. can represent taxpayers. AI cannot hold a license. AI cannot be sanctioned for violating Circular 230. AI cannot sign a return as a paid preparer. The licensing layer is real and durable.

Liability. When a return is wrong, someone is responsible. The IRS doesn't audit ChatGPT. It audits the preparer who signed the return. The liability layer means human judgment is required at the point of filing, regardless of what software assisted in preparation.

Face-to-face trust. People with complex tax situations. business owners, gig workers, expats, people under audit. want to sit across from a human being who can explain what's happening. Software cannot do this. The more AI automates routine returns, the more the remaining human-preparer clients are the ones who most need a human.

Supply collapse. The demographic cliff is unaffected by any AI stock price. The average CPA is 51. Roughly 15,000 to 20,000 preparers retire each year. New entrants: 8,000 to 12,000. Net annual loss of 5,000 to 10,000. AI hype or no AI hype, the people doing this work are aging out faster than they're being replaced.

None of these things change when a semiconductor stock drops. The protections are structural, not narrative.

The fear is what was hurting the profession

The real damage from the AI hype cycle to tax preparation wasn't automation. It was the story.

When every media headline says AI will eliminate tax preparers, three things happen. Potential entrants choose other careers. Current preparers delay investing in their practices. Clients question whether they need a human preparer at all. The narrative creates real behavioral effects even if the underlying threat never materializes.

But the same narrative also creates a moat. The people who are scared away by AI headlines leave the field to the people who understand what the data actually says. Fear is a barrier to entry. When the fear deflates, the barrier lowers. but the structural protections remain.

The bursting bubble means the fear narrative loses its power source. When Nvidia's stock drops 5% in a day, "AI will replace all white-collar workers" sounds less like a certainty and more like what it always was: a projection, not a fact.

What actually changes for EA candidates

The 45,000-plus tech workers who've been laid off in 2026 aren't going back to the same jobs. Some will reskill. Some will start businesses. Some will look for professions that AI cannot easily automate and that have clear credential paths.

Tax preparation checks both boxes. The EA credential requires three exams, no degree, and costs under $1,000 total. The demand for preparers is growing as the supply shrinks. The work cannot be offshored to a language model.

A laid-off product manager or junior developer who spends six months studying for the EA exam can be a licensed tax professional by next tax season. The credential doesn't care about their previous job title. It cares about three test scores.

During a tech selloff, a federal license with liability protection starts looking better than a job that can be eliminated in a restructuring email. The same forces that crashed the NASDAQ this week are the forces that make the EA path more attractive, not less.

The KB's three scenarios

I've written before about the AI labor curves and what they mean for tax prep. But there's a simpler framing that might be more useful right now.

There are three possible futures for AI. In all three, the tax preparation position holds.

Scenario A: AI keeps improving. Better tools for preparers. More efficiency. More complex returns as gig work expands. The position benefits from the accelerant.

Scenario B: Bubble bursts, AI plateaus. This is what's happening now. The fear narrative collapses. Clients stop asking "why not use ChatGPT." The supply-demand gap remains. The position actually gets stronger because the irrational fear barrier drops.

Scenario C: AI was always overblown. The tools that worked still work. The office that existed before LLMs still exists. The 10-plus years of client relationships still matter. The position was never dependent on AI anyway.

There is no scenario where AI makes the EA credential useless and tax preparers obsolete. That scenario was always a narrative, not a forecast. The bubble bursting doesn't create that scenario. It kills the narrative that claimed it was inevitable.

If you've been putting off the EA exam because you weren't sure AI would leave anything for human preparers to do, the market just gave you an answer.

Start studying. free EA practice questions →

Keep building this concept · 3 related guides

EA Exam Prep in the Bay Area: Tech Equity, AI Displacement, and California's 13.3% Tax BracketThe Bay Area has the highest concentration of tech wealth in the world, California's 13.3% top bracket, and a layoff cycle that's been turning W-2 workers into 1099 consultants since 2023. The EA path fits this market perfectly.Becoming an Enrolled Agent From India: The 2026 Path, Costs, and Real Career OutlookIndia produces more EA candidates than any country outside the US. Remote testing is now available. Over 1,500 US tax outsourcing firms operate in India. Here's the real path, what it costs, and what the job market actually looks like.A Denver Tax Career Without an Accounting Degree: The Enrolled Agent PathDenver has a large accounting workforce, national replacement demand remains strong, and the IRS offers a federal tax credential that does not require an accounting degree. Here is what the evidence supports—and what it does not.

Find your weak topics →Open the 50-unit path →