The short answer
In one paragraph
A small accounting or bookkeeping firm should automate document collection and transaction categorisation first. These tasks repeat every month, follow clear rules and are easy to check, so the time saved shows up quickly and errors are caught at review. Next, use AI to draft client emails, meeting notes and first-pass research. Keep the judgement calls, the sign-off and the client conversations with your people. One caution: client tax data should only go into vetted, secure business tools, never a free public chatbot. Guidance from the IRS in June 2026 says the practitioner must check every AI-generated fact, figure and citation before it reaches a client.
By Ian Luca Calcavechia, AI consultant, Kraków · Last updated 2026-11-13
AI in accounting and bookkeeping firms: where things stand in 2026
- 41% of tax and accounting firms said they use AI in 2025, up from 9% in 2024, in Wolters Kluwer's Future Ready Accountant study of more than 2,700 professionals. 77% plan to increase AI investment over the next three years. (Accountancy Today, October 2025)
- Enterprise use of generative AI at tax firms rose to 21% in 2025 from 8% in 2024. 77% of tax firm clients want their firm to use it, yet 59% do not know whether it does. (Thomson Reuters, April 2025)
- Across legal, tax and accounting and other professional services, organisation-wide AI use reached 40% in 2026, up from 22% in 2025. Only 18% of respondents said their organisation tracks the return on its AI tools. (Thomson Reuters Institute, February 2026)
- CPA.com's 2025 AI in Accounting report says some firms have automated more than 80% of individual tax return preparation, and reports gains of up to 70% less time on manual tasks. The report adds that most firms cannot yet measure the return. (Accounting Today, June 2025)
- Staff at small firms (5 to 19 employees) were less positive about AI than staff at large firms, 54% against 76%. Where the firm had an AI policy, 84% of staff were optimistic, against 44% where it had none. (CFO Brew, October 2025)
Six uses that pay off first
Automated requests and reminders that chase missing bank statements, receipts and tax documents, and file what arrives in the right client folder. Staff only step in when a client stays silent.
AI suggests categories and matches bank lines for reconciliation, and a bookkeeper reviews the exceptions. The routine lines no longer need a human touch every month.
New emails are sorted by client and urgency, and routine replies are drafted for a person to approve. Questions that need judgement go straight to the right accountant.
A web form collects the facts you need before a first call and sends them to your practice management tool. Poor-fit prospects are screened out before anyone books time.
Client calls are summarised into action items and a draft follow-up email. The accountant checks the summary before it is saved or sent.
First drafts of tax research memos, month-end commentary and variance explanations. Every citation and number is checked against the source before the client sees it.
How leading companies use it
Built a prospect intake form with Lovable, embedded in its website with help from Claude, that collects industry, headcount, revenue and reasons for switching, with CAPTCHA and rate limiting, and sends answers into ClickUp. Conversion rate rose about 25%, and the partners spend less time on calls with prospects who do not fit. (Journal of Accountancy, August 2026)
The COO used Claude Code to build an app connected to Karbon that asks clients for satisfaction feedback after major projects or monthly, replacing costly satisfaction survey programs. The head of tax, with no tech background, built a prototype onboarding and tax organizer. The feedback app took about four or five hours to build; the main running cost is a Claude team account. (Journal of Accountancy, August 2026)
Set up a Microsoft Power Automate workflow for nonprofit clients: an invoice sent to a dedicated inbox triggers an Adobe Sign approval, then a QuickBooks bill is created with the signed approval attached. No cost beyond existing Microsoft 365 and Adobe subscriptions, no new logins for clients, and a complete audit trail. (Journal of Accountancy, August 2026)
Moved transaction categorisation, reconciliations and reporting to an AI-based accounting platform, connected to an internal system the founder built. According to the founder, bookkeeping time fell by about 75%, and three years of historical clean-up took two days. (Accounting Today, April 2026)
Launched KPMG Workbench, a single AI platform with 50 AI agents and assistants and nearly 1,000 more in development, behind its tax, advisory and audit platforms. (KPMG, June 2025)
What’s working now: the boldest AI already in use
The boldest AI that is actually running, not concept videos: named organisations, dated, with the source. Figures marked company- or vendor-reported have not been independently audited.
Basis · USA
AI agents connect to a firm's ledgers and ERPs, take in client documents and run multi-step processes such as reconciliations, journal entries and month-end close over several hours, returning finished work for human review. Result: Raised a US$100 million Series B at a US$1.15 billion valuation in February 2026; the company reports efficiency gains of 20% to 50% across practices, with no third-party verification (company-reported, via secondary coverage). (Accountio)
Copy it small: An agent that categorises transactions, reconciles bank feeds and drafts the month-end checklist and variance notes; nothing posts without an accountant's approval.
Main risk: A consolidating market: the AI bookkeeping platform Botkeeper shut down in February 2026 and Xendoo bought its platform, so choose vendors you can leave. (source)
AI playbook by budget: 15 moves for accounting and bookkeeping firms
Fifteen moves for accounting and bookkeeping firms, from what any business can switch on this week to the bets that need serious money and development. Costs and timelines are typical ranges (estimates), not quotes. Named results link to their source; company and vendor figures are labelled. Every move keeps a person approving anything that goes to a customer, a payment or a legal or clinical decision.
Start this week
- Receipt captureTools: Dext, Hubdoc, QuickBooks or Xero AI · Measure: hours per client
- Client email draftsMeasure: response time
- Meeting notes into filesTools: Fathom · Measure: follow-ups done
- Tax-research draftsVerified by a CPA.Measure: research hours
- Engagement-letter templatesMeasure: onboarding time
Invest for a few months
- AI categorisation and reconciliationTools: Booke.ai, Digits, Basis · Measure: clients per bookkeeper
- Practice-management automationTools: Karbon, TaxDome · Measure: days to collect documents
- Close automation with variance commentaryMeasure: close days
- Advisory dashboardsMeasure: advisory revenue
- Document-to-workpaper tax preparationMeasure: returns per preparer
Bet big
- AI-native fixed-fee firmMeasure: revenue per employee
- Cross-client benchmarking dataMeasure: advisory attach rate
- Fractional CFO service with AI forecastingMeasure: recurring revenue
- Acquisitions on an AI back officeMeasure: margin uplift
- Niche vertical softwareMeasure: ARR
What not to do
- Pasting client tax data into public chatbots. Guidance from the IRS says client data should be handled only with secure, enterprise-approved AI and never uploaded to unsecured sites. Unauthorised use or disclosure of return information can bring civil and criminal preparer penalties. (source)
- Sending AI output without a real review. Deloitte agreed to repay part of an AU$440,000 Australian government contract after a report it produced with help from a GPT-4o tool chain contained fake citations and a made-up court quote. (source)
- Billing for hours the AI did. The IRS warns that billing for work not actually performed, or double billing for AI-assisted tasks, may breach Circular 230's fee rules, and that AI savings should be credited to the client. (source)
- Using return data for new services without consent. Using tax return information to market other services, or sharing it with a provider that makes decisions on the return, generally needs the client's consent, obtained in advance, under Section 7216. (source)
- Skipping the written AI policy. Staff at firms with an AI policy were far more positive about AI than staff at firms without one. A one-page policy on approved tools, client data and review is enough to start. (source)
Rules to know
- The IRS Office of Professional Responsibility's Introductory Guidelines for Responsible AI Use in Federal Tax Practice apply existing Circular 230 duties to AI: check AI-generated facts, citations and calculations (section 10.22), understand the tools you use (10.35), protect return information (sections 6713 and 7216), bill fairly for AI-assisted work (10.27) and put firm AI policies and training in place (10.36). (US IRS Office of Professional Responsibility, June 2026)
- Under Section 7216, a preparer who knowingly or recklessly discloses or uses tax return information faces a fine of up to $1,000, up to one year in prison, or both. Disclosures for purposes beyond preparing the return generally need the taxpayer's consent obtained before the disclosure. (The Tax Adviser, January 2024)
- Federal law (the Gramm-Leach-Bliley Act and the FTC Safeguards Rule) requires tax and accounting professionals to keep a Written Information Security Plan and to report certain security events affecting 500 or more people, generally within 30 days. Publication 5708 from the IRS gives a template, and firms must choose service providers that can maintain appropriate safeguards, which applies to any AI vendor you add. (US IRS, August 2026)
Practical guidance from an AI consultant, not legal advice; check each build with your adviser or compliance team.
How to start in 30 days
- Week 1: List the ten most repetitive tasks in the firm and time them for a week. Write a one-page AI policy: approved tools, no client data in public chatbots, and a person reviews every output.
- Week 2: Pick one workflow, usually document chasing or prospect intake, and build it on business-grade tools you already pay for. Add any new vendor to your Written Information Security Plan.
- Week 3: Run it with five to ten friendly clients. Review every output, log errors and measure the hours saved against week 1.
- Week 4: Decide to keep, fix or drop it. Check whether any new use of return data needs Section 7216 consent, then train the whole team on the workflow.
- Week 5: Add the second workflow, such as transaction categorisation with exception review, and repeat the same measure-and-review cycle.
Tools in use in this sector
QuickBooks Online · Xero · Karbon · Blue J · Microsoft Copilot and Power Automate · ChatGPT Enterprise · Claude · Zapier. The right one is usually the one the team already uses; automations in n8n, Make or Zapier connect them.
Work with an AI consultant for accounting and bookkeeping firms
Set up the first moves for accounting and bookkeeping firms in days, or build the connected system (n8n, Make or Zapier agents, CRM, integrations) over a few months.
AI automation on Fiverr →An AI roadmap for your business: which moves pay off first, tools, costs, risks, and an EU AI Act check.
AI roadmap on Fiverr →Send the process, the vendor proposal or the AI tool you are considering; get a written second opinion on fit, cost, risks and what to ask.
Quick scan on Fiverr →For a Tier 3 bet: a PRD, roadmap and backlog, and someone running the build with your developers and vendors.
Product owner on Fiverr →Ian Luca Calcavechia is an independent AI consultant in Kraków working in English, Spanish, French and Russian, AI product owner and delivery lead for a US healthcare company since 2022, freelancing since 2019 with 100+ Fiverr orders. All services and prices · All 30 industries
Who wrote this
Independent AI consultant who builds AI roadmaps, AI automations and agents (n8n, Make, Zapier) and custom CRMs for small and mid-sized companies in the EU, the US and Latin America. Since 2022 AI product owner and delivery lead for a US healthcare company, where he rebuilt a multi-party operation as a custom CRM with five portals and an AI layer that turns calls, texts and emails into case notes. Services and prices · All industries
Get it done
Start with a fixed package
AI automation
One workflow for accounting and bookkeeping firms, with an AI step and a walkthrough video
Start on Fiverr →AI roadmap
Where AI pays off first in accounting and bookkeeping firms, tools, costs and risks
Start on Fiverr →Questions
Frequently asked
How much does AI cost a small accounting firm?
Often less than expected, because much of it runs on tools the firm already pays for. A sole practitioner in Florida built an invoice approval workflow with no cost beyond Microsoft 365 and Adobe, and an eight-person firm in Arizona says its main AI cost is one team subscription. The bigger cost is staff time to set up and review the first workflows.
Can we put client tax data into ChatGPT?
Not into a free public version. Guidance from the IRS says client data should be handled only with secure, enterprise-approved AI tools, and Section 7216 limits how return information can be shared or used. Use business accounts with data protection terms and record the vendor in your security plan.
Will AI replace our bookkeepers?
The job changes more than it disappears. AI takes over categorisation, matching and first drafts, while people handle exceptions, client relationships and judgement. CPA.com's 2025 report says AI fluency is now expected even in entry-level roles, and junior staff are judged more on analysis than on task completion.
Which AI tools do accounting firms actually use?
Common choices are the AI features inside QuickBooks Online and Xero, practice management tools like Karbon, research tools like Blue J, and general assistants such as ChatGPT Enterprise, Claude or Microsoft Copilot. Automation tools such as Power Automate or Zapier connect them to email and client portals.
Where should a small firm start?
Start with the task that repeats most and is easiest to check, which for most firms is chasing client documents or categorising transactions. Write a short AI policy first, run one workflow for a month with a few clients, and measure the hours saved before adding the next.
Who can implement AI for accounting and bookkeeping firms?
Ian Luca Calcavechia, an independent AI consultant, implements AI for accounting and bookkeeping firms: quick setups that take days, connected systems with AI agents, CRMs and integrations built over a few months, and product ownership for bigger bets. Fixed packages start on Fiverr (fiverr.com/ianlucac).
Can someone review an AI vendor's proposal or a specific case for accounting and bookkeeping firms?
Yes. Ian Luca Calcavechia reviews a specific process, vendor proposal or AI tool and gives a written second opinion on fit, cost, risks and the questions to ask, starting with the AI quick scan package on Fiverr.
What is the cheapest AI to start with in accounting and bookkeeping firms?
The moves in the Start this week column: AI answering and booking for calls, chats and emails, and AI drafts of documents the business already writes, with a person approving what goes out. They usually cost under US$100 a month (estimate).
More guides
- AI for farms and agri-food exporters: faster quotes, cleaner paperwork, smarter fields
- AI for car dealerships: answer every lead, book more service, keep the deal human
- AI for construction companies: less paperwork, faster reports, better follow-up on bids
- AI for credit unions and community banks: better member service, safely governed
- AI for dental clinics: fewer missed calls, faster insurance checks, more chair time
- AI for online stores: better product data, faster support, fewer returns
- How much does AI automation cost for a small business in 2026
- How to hire an AI automation consultant: 10 questions to ask before you pay
- AI receptionist for a small business: what it costs in 2026 and what to check first
- Website AI chatbot for a small business: what it costs in 2026 and what to check first
- AI use policy for a small company: a one-page template and how to roll it out
- AI for customer emails in a small business: tools, prices and what to keep with a person
- AI note-takers in meetings: consent, retention and the rules for a small company
- All industries