The IRS Already Knows What Your Clients 'Should' Be Spending — And Flags the Outliers

April 9, 20266 min readBy The Crossing Report

Published: April 9, 2026 | By: The Crossing Report


Summary

The IRS now operates 129 AI use cases — up from 54 in 2024. One of the most consequential for your clients: an AI system that compares every Schedule C line item against model-derived benchmarks for that specific industry and zip code. A consultant with 80% travel expenses when the industry norm is 20% is flagged automatically. The system runs approximately six times per tax year. For accounting firms with small business clients, this changes the advisory conversation: expense ratio review is now a proactive client protection service, not just an audit defense.


The IRS AI System That's Already Benchmarking Your Clients

Most accounting firm owners know the IRS uses technology to match returns against third-party data — W-2s, 1099s, K-1s. That part is old news.

Get the full picture. Go premium.

Weekly intelligence briefings, deeper analysis, and direct access to the full archive.

What's changed in 2026 is the sophistication of what happens before a return is flagged for human review. The IRS now operates 129 AI use cases. One of the most operationally significant: a system that builds a statistical model of what businesses in each industry and zip code "should" spend — and flags returns where the reality doesn't match the model.

For a Schedule C filer, this means the IRS AI isn't just checking whether reported income matches a 1099. It's comparing every major expense category against what the system expects for a business of that type, in that location, at that revenue level.

A photographer claiming 60% of revenue in equipment depreciation? Probably fine — that's consistent with the industry profile. A management consultant claiming 80% travel expenses when the industry norm is 20%? Flagged.

The system runs approximately six times per tax year. It learns continuously. And it's now operating at scale after the IRS workforce was reduced by 25% in 2024, making AI-assisted enforcement the primary infrastructure, not the supplement.


What the AI Is Actually Looking At

The IRS AI benchmarks against several categories for Schedule C businesses:

Expense ratio by category. Home office, vehicle, meals and entertainment, travel, contractor payments, equipment — each compared against the distribution for the same NAICS or SIC industry code. Outliers above approximately the 90th percentile for any category increase the risk score.

Year-over-year patterns. A business that shows consistent revenue growth but suddenly claims a 40% increase in supplies expenses — without a corresponding increase in revenue — is flagged as an anomaly even if the absolute amount looks reasonable.

Third-party mismatches. Income reported on a Schedule C is cross-referenced against 1099s, bank-reported interest, and other third-party filings. Gaps — even small ones — add to the risk profile.

Consecutive-year loss patterns. Businesses claiming losses in three or more consecutive years on Schedule C receive elevated attention. The system distinguishes between businesses in a legitimate growth phase (revenue trending up, expenses trending toward profitability) and businesses that appear to be generating paper losses from personal expense reclassification.

Geographic norms. The same expense ratio might be unremarkable in San Francisco and flagged in rural Ohio. The model accounts for local cost-of-business variation.


What This Means for Your Practice

This is a client advisory story with a specific revenue angle.

The accounting firm that proactively tells clients "here's how your expense ratios compare to the IRS benchmark for your industry" is delivering a protection that was theoretically available before and is now table-stakes for small business advisory. You can do this work. Your clients need it. Most of them don't know this benchmarking system exists.

Three concrete actions:

1. Build a benchmark checklist by client industry. For each major Schedule C client segment you serve — independent consultants, real estate professionals, medical practitioners, retailers — pull the IRS industry average expense ratios. The IRS publishes statistics through its SOI (Statistics of Income) reports, and several accounting software platforms are building this into their advisory modules. Know the norms before your clients file.

2. Flag deviations before filing, not after. When a client's expense ratio in any category deviates from the industry benchmark by more than 20–25%, document the business rationale in writing. If the deviation is legitimate (a photographer with genuine equipment costs), the client should have receipts, depreciation schedules, and a one-page narrative explaining why their expenses look the way they do. That documentation is the audit defense. The time to build it is at filing preparation, not when the IRS notice arrives.

3. Frame the conversation as competitive advantage. Most small business owners have no idea the IRS AI benchmarks their expenses against industry norms. Telling them — before filing — positions your firm as the advisor who protects them proactively. That framing is more valuable than explaining it as a compliance exercise. "Here's what the IRS AI will compare you against, here's where you stand, and here's what we're documenting to protect you" is a differentiated advisory conversation.


The Revenue Angle for Accounting Firms

Expense benchmarking and pre-filing audit risk review are advisory service lines, not just tax preparation steps. If you are currently offering this analysis as part of a standard tax prep engagement without pricing it separately, you have a service unbundling opportunity.

What does "pre-filing IRS AI risk review" look like as a standalone or premium advisory service?

  • One-hour engagement for each Schedule C client, reviewed against the IRS benchmark model for their industry
  • Written summary of deviation flags and documentation requirements
  • Annual audit risk score for the client's profile

For clients who run significant Schedule C businesses — contractors, consultants, real estate professionals — this is a service with tangible protective value. The IRS AI runs six times a year. Your advisory should run at least once, before filing.


The Broader Context

The IRS AI expansion isn't slowing. 129 use cases from 54 in two years reflects an agency that has structurally reorganized around automated enforcement. Large partnerships ($10M+ assets) face heightened scrutiny this year. Self-employed taxpayers and small business owners with meaningful expense deductions are the next tier.

The accounting firms that understand the mechanics of how AI audit selection works — and build advisory processes around those mechanics — are the firms that will justify their fees to clients who are wondering whether TurboTax is sufficient. The answer is: TurboTax can file the return. It cannot tell the client that their entertainment expense ratio is in the 94th percentile for their industry and here's how to document the exceptions.

That's your value. Build the process around it.


Action step this week: Pull the IRS Statistics of Income tables for the top two or three industries your Schedule C clients operate in. Note the average expense ratios for the major categories. Compare those norms to your last three clients in each industry. If anyone deviates by more than 20% in any category — do you have documentation for why? If not, that's the conversation to have before their next return is filed.

Sources: Klawson Law — IRS Using AI to Target More Audits in 2026 | Capitol Technology University — Audited by an Algorithm | EisnerAmper — AI at the IRS

Frequently Asked Questions

How does the IRS use AI to flag Schedule C expenses?

The IRS AI system compares each Schedule C line item against a model-derived benchmark of what similar businesses in the same industry and zip code typically spend. A consultant claiming 80% of revenue in travel expenses when the industry average is 20% is automatically flagged. The system benchmarks against specific industry codes (NAICS/SIC) and geographic location, runs approximately six times per tax year, and is continuously updated as the IRS collects more data. The result is a dynamic risk score for each return — not a binary flag — that determines audit probability.

Which taxpayers are at highest risk from IRS AI audit targeting?

Three groups face the most elevated audit risk from IRS AI in 2026: large partnerships (over $10M in assets), self-employed taxpayers filing Schedule C, and small business owners with significant business expense deductions. The AI system is particularly effective at identifying expense patterns that deviate from industry norms — large home office deductions relative to revenue, high vehicle expense ratios, or meals and entertainment that exceed sector benchmarks. Clients with multiple consecutive years of losses on a business Schedule C are also flagged as a pattern.

What should accounting firms do to protect Schedule C clients from IRS AI audits?

Three specific actions. First, build an expense ratio benchmark checklist for each client's industry code — compare their expense categories against IRS industry averages and flag anything that deviates by more than 20%. Second, when a legitimate expense is an outlier (e.g., a photographer who genuinely spends 60% on equipment), document the business rationale in the client file and advise the client to retain supporting records proactively. Third, run this analysis before filing, not after — anomaly detection is more useful as a pre-filing advisory than as audit defense.

How many AI use cases does the IRS now operate?

As of 2026, the IRS operates 129 AI use cases — up from 54 in 2024. The rapid expansion was accelerated by a 25% IRS workforce reduction, which made AI the core enforcement infrastructure rather than a supplemental tool. The AI systems handle audit selection, document processing, identity verification, and taxpayer risk scoring. The agency runs AI-assisted matching against third-party reporting (W-2s, 1099s) on an ongoing basis, not just at filing season.

What does the IRS AI system look at beyond expense ratios?

In addition to Schedule C expense ratios, the IRS AI monitors: year-over-year inconsistencies in reported income or expenses, mismatches between reported income and third-party forms (W-2, 1099, K-1), deductions that are unusually large relative to gross income for the industry, and cash-intensive business patterns that diverge from industry norms. The system also analyzes employer payroll tax patterns, estimated tax payment history, and prior-year audit outcomes when scoring current returns.

Get the weekly briefing

AI adoption intelligence for accounting, law, and consulting firms. Free to start.

Free weekly digest. No spam. Unsubscribe anytime.

Related Reading

This is the kind of intelligence premium subscribers get every week.

Deep analysis, cross-sector patterns, and the frameworks that help professional services firms make the crossing.