Clients don't want the trial balance. They want to know why they owe more this year and what to do about it. That translation from numbers to plain English is where accountants lose the most unbilled time, and it's exactly what AI is built for.
The firms getting real value aren't using AI to file returns. They're using it to draft the explanations, client emails, and categorization work that surround the actual accounting, so more hours go to analysis clients will actually pay for.
Ways accountants use AI
Explain a tax bill or financial statement in plain English
Ask AI to draft a plain-language explanation from verified figures. The accountant checks the numbers and reasoning before sharing the explanation with the client.
Draft categorization for messy transactions
Feed AI a batch of ambiguous transactions and it suggests categorization with reasoning, so review takes minutes instead of line-by-line guessing.
Write engagement letters and scope emails fast
AI drafts the engagement letter or scope-change email from the specifics of the client and service, so onboarding a new client doesn't mean starting the document from zero.
Turn an IRS or state notice into a client explanation
Drop in the notice and AI drafts a plain-English summary of what it means and what the client needs to do, cutting the panicked-client phone call down to a two-minute read.
Draft monthly close summaries for bookkeeping clients
AI turns the month's financials into a short narrative clients actually read, the kind that makes them feel the retainer is worth it instead of just a PDF nobody opens.
Prep a first draft of a tax organizer follow-up
AI drafts personalized follow-up messages to clients with missing documents, chasing the same handful of forms without you writing the reminder five hundred times each January.
Where to start this week
- The client explainer: take one client's tax bill or P&L and have AI draft the plain-English explanation you'd normally write yourself.
- The transaction batch: hand AI a month of uncategorized transactions and review its suggested categorization.
- The engagement letter: build a reusable prompt that drafts a scope-specific engagement letter from a few client details.
Bring one real task. Find a first build you can test.
Find my first build ↗Frequently asked questions
What is the best AI tool for accountants?
A general AI assistant with strong prompts handles client communication, categorization, and letters without another subscription. Bookkeeping software's built-in AI features are worth using for pure data-entry automation, but drafting still benefits from a strong general model.
Can AI do my clients' taxes?
No, and it shouldn't try. AI drafts explanations, categorizations, and client communication, but the actual return, the judgment calls, and the sign-off stay with the accountant, who carries the liability.
Will AI replace accountants?
These examples focus on draft preparation and review. They do not predict how jobs will change. Accountants remain responsible for the work they choose to use and the decisions that follow.