60% of tax firms now use AI for research — and billing models are cracking

No tool in this week's batch meets the accessibility bar — publicly available today, transparent pricing, no sales call required, usable by a solo practitioner. Back next issue.

Tax research AI doubles — billing models follow — 60% of US tax practitioners now use AI for research, up from 33% last year, with many exploring alternatives to the billable hour, per CPA.com's survey of 1,000 US accountants. Why it matters: If your pricing model still assumes research takes the same time it did in 2024, your rates are already out of step with what AI-equipped competitors are charging. [Read more →]

Agentic AI — owns the workflow, not just the task — AI agents that receive a goal and execute a full sequence of steps without prompting are already entering tax practice, replacing tools that only respond when asked. Why it matters: Firms that treat this as a better search engine will be caught off guard when agents start handling entire engagements autonomously. [Read more →]

Clients filing with AI — nobody reviewing the work — Freelancers, landlords, and sole traders are using ChatGPT and Claude to produce tax summaries they file directly, with output polished enough that they don't know what they're getting wrong. Why it matters: The profession has spent two years debating whether AI replaces accountants; the more immediate problem is that clients are already using it unsupervised, and some of that work will surface at audit. [Read more →]

R&D credits — AI can file, not qualify — AI platforms can streamline R&D credit filings but cannot determine whether a project meets the four-part qualified research test, and the signing CPA remains fully liable under Circular 230 regardless of which platform generated the work. Why it matters: Firms receiving client-generated R&D filings from third-party AI platforms need a clear policy on where their review obligation begins before the IRS asks. [Read more →]

MindBridge expands — full-population audit analysis now — MindBridge added subledger analysis, a Monetary Flow Dashboard, and enhanced transaction-level risk scoring, shifting audit workflows from statistical sampling to full-population review. Why it matters: Audit teams carrying more clients per staff member can now flag risk automatically across entire ledgers rather than defending a sample. [Read more →]

AI recruiting risk — CPA firms face resume fraud — AI-polished resumes are clearing keyword screens at firms where candidates will handle sensitive client tax files, audit workpapers, and financial records, with credential claims harder to verify than ever. Why it matters: Structured skills testing before extending an offer is no longer optional for firms that can't afford a bad hire with access to client data. [Read more →]

AI is moving fast and most people are falling behind. 

The Rundown AI keeps you ahead of the curve. 

It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.

Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses — tailored to your needs.

Sign up to start learning.

Run one verifiable question through your AI research tool this week — something you already know the answer to — and check whether it cites current law or a pre-OBBBA version. The adoption numbers above assume the tool is returning accurate output; not all of them are.

State AI taxes: State lawmakers are introducing tax legislation targeting AI services, with Indiana and Illinois ruling chatbot services non-taxable while Michigan and others signal different approaches — all before any federal framework exists. Why it matters: CPAs advising clients who pay for AI tools or sell AI-adjacent services need to track how their state classifies that spending before the rules solidify. [Read more →]

Synthetic ID fraud: AI-powered synthetic identity fraud — real Social Security numbers combined with fabricated identities — is projected to generate $3.1 billion in US unsecured credit losses in 2026, up from $1.8 billion in 2020, growing at 16% annually. Why it matters: Small business clients applying for credit or managing vendor onboarding are in the blast radius, and their CPA is often the first call when something goes wrong. [Read more →]

Generic AI and the audit trail: Three structural gaps — traceability, repeatability, and defensibility — consistently prevent generic AI tools from being used safely inside financial reporting workflows, according to an Accounting Today analysis of how finance leaders are actually deploying AI. Why it matters: Every number in accounting must survive scrutiny; firms running ChatGPT or Copilot inside their close process should confirm their outputs meet that bar before a client or regulator asks. [Read more →]

AI burnout signal: A CPA who nearly left the profession describes how delegating data entry and deadline grind to AI gave back the client-facing work that made the job meaningful — at a moment when 71% of remaining finance professionals are actively considering leaving. Why it matters: Retention is now a legitimate argument for AI adoption inside firms, alongside the efficiency case. [Read more →]

The signal this week: AI has stopped assisting and started owning — the full return, the full audit, the full research workflow. The firms that treat that as an upgrade will be ready. The ones that treat it as a better search engine won't.

—Alex

Implementing usage-based pricing successfully requires more than just a pricing strategy.

Download this guide for practical advice and best practices when considering usage-based pricing. 

👉 Get your guide

📬 Know an accountant drowning in manual bookkeeping? Forward this.