Debit AITHE ROLL-UP TRACKER
Subscribe free →

DEBIT AI · UPDATED EVERY MONDAY

The Roll-Up Tracker

Who's buying accounting firms — and what those firms are buying next. Private equity, mergers, multiples, and the AI actually being deployed inside the profession.

⚠️ Deal terms and AI-investment figures are as publicly reported, and are frequently partial or re-labeled. Verify before citing. A field guide, not investment advice.

Get it every Monday Jump to the deals ↓

Why two tracks

Step 1 PE buys a firm at a low-to-mid-teens EBITDA multiple.
Step 2 That price only pencils out if margins expand.
Step 3 Firms are labor-heavy and lightly digitized.
Step 4 So the fastest margin lever is AI and automation.

The deal predicts the spend. Which is why this tracker follows both — who buys the firms (Track A), and what those firms then buy (Track B). Track B is often the more useful half: when a firm of consequence commits real money and real liability to a tool, that is the best available evidence it has crossed from demo to production.

📌 This week

WEEK OF AUGUST 24, 2026

Track A · Who's buying firms

Private-equity investments, firm mergers, and platform tuck-ins. Nearly all use an alternative practice structure (APS): the firm splits into a CPA-controlled attest entity and a non-attest advisory entity, and the outside capital goes into the advisory side. That split is the regulatory key that makes the entire wave possible.

Aug 21
2026

Grant Thornton Advisors → Grant Thornton Australia (completed)

Terms not disclosed — no consideration, no structure, no multiple

Sydney-based Grant Thornton Australia has completed its transaction to join the multinational platform of Grant Thornton Advisors (New Mountain Capital-backed; FY25 net revenue $2.46 billion), adding 1,500 professionals and six offices. The Australian firm reported annual revenue of about $282 million when the deal was announced in April. It follows Grant Thornton New Zealand, which joined in late 2025 with 300+ professionals. Said Jahani, CEO of GT Australia, has been named CEO of the Asia-Pacific region, joining the platform's executive committee and reporting to CEO Jim Peko. GT Australia remains a member of Grant Thornton International Limited.

→ Sponsor-backed consolidation has gone cross-border, and the logic is the same one behind CBIZ. Read the two together: a $5 billion US acquisition agreed in July, a 1,500-person Asia-Pacific firm closed in August, both under a platform that committed $1 billion to AI in September 2025. That spend needs a denominator, and the platform is assembling one — post-CBIZ it expects 34,500+ professionals across 20+ countries. The question for a reader running a domestic mid-market practice isn't whether to compete with this; it's what a competitor with that cost base can charge. ⚠️ IPA states the $282M in USD; the firm reports in AUD and no conversion basis is given — treat as approximate.
Aug 11 & 19
2026

CRI → DMR CPAs, Trace Advisory

Both undisclosed

Carr, Riggs & Ingram (Top 25; Centerbridge Partners and Bessemer Venture Partners-backed since late 2024) closed two deals in nine days: DMR CPAs of Fort Worth, Tex. (founded 1958; tax, audit, accounting, payroll), announced Aug 11; and Trace Advisory of Ridgeland, Miss. (government accounting, auditing and compliance), announced Aug 19. These are CRI's third and fourth M&A deals of 2026, and it has made 13 transactions since taking private equity in late 2024.

→ Note what happens to the target. Trace Advisory doesn't get absorbed — it becomes CRI Government Solutions, a named practice line. That's the same shape as UHY buying Michigan CFO Associates for its CAS capability a week earlier: the platform is increasingly buying a capability it can brand, not a postcode it can staff. Capability buys are the ones to watch, because they are where firm-side automation lands first and where a specialist practice can still command a premium. ⚠️ The "13 transactions" count appears to include these two — they are not the 13th and 14th.
Aug 19
2026

BST & Co. + Gallagher Flynn — no sponsor

Undisclosed · merger of equals, no outside capital

Latham, N.Y.-based BST & Co. CPAs (IPA Top 300) and South Burlington, Vt.-based Gallagher, Flynn & Co. have combined to create BST Gallagher Flynn. No private-equity capital is involved.

→ The control case, and worth logging precisely because it's unremarkable. Financial acquirers are now 54.8% of accounting-firm M&A (Capstone, July) — up from 38.9% a year earlier. Against that, a straightforward two-firm combination with no sponsor shows what firms of this size do when they are not solving a partner-buyout problem with someone else's balance sheet. Keep an eye on the count: the interesting question in twelve months is whether deals like this still exist at this scale, or whether the non-sponsor route has closed.
Aug 19
2026

IPA data → growth 12.8%, organic 7.0%

Market data — not a deal · INSIDE Public Accounting, 2026 IPA 500 analysis

Among the IPA 100: total net revenue growth reached 12.8%, up from 10.4% a year earlier. Organic growth moved the other way, falling from 7.8% to 7.0% — described by IPA as its slowest pace in five years. The IPA 100 completed 211 mergers during the most recent survey year. The 10 fastest organic growers averaged 15.6%, more than double the IPA 100 average.

→ This is the number that frames the entire tracker. Headline growth is accelerating while internally generated growth decelerates; the gap between them is bought. And bought revenue arrives without bought margin — margin has to be manufactured afterwards, and in a labour-heavy, lightly digitised business the only lever that works at scale is automation. Which is to say: the thinner organic growth gets, the more a sponsor's return depends on Track B actually delivering. It also sharpens Capstone's July finding that AI readiness is now priced into the deal rather than discovered after it. Watch the organic-versus-total gap in the 2027 data — it is the cleanest single indicator of whether the roll-up is creating value or merely aggregating it.
Aug 7
2026

KKR → Crowe Advisory LLC (closed)

Nothing disclosed at close — no value, no stake size, no multiple (the ~$3B / majority figures reported in June remain unverified)

Funds managed by KKR have closed on their equity investment in Crowe Advisory LLC, the non-attest entity of Crowe — the #12 US firm, ~$1.39B revenue, 5,600+ staff, and one of the last large holdouts. Announced June 11, closed August 7. Quotes from Crowe CEO Steven Strammello and KKR partner Chris Harrington. Both entities remain members of the Crowe Global network.

→ The close release answers the APS question the announcement left open — and the answer matters. Verbatim: "Crowe Advisory LLC will provide tax, advisory and other non-attest services, while Crowe LLP will remain a licensed CPA firm and continue to provide attest services, including audits and reviews." The split is not audit-versus-consulting. It's audit-versus-everything-else, and for most mid-market firms tax is the profit engine. If you're reading your own term sheet, this is the line to find first: which entity does tax live in?
Aug 11–14
2026

Four tuck-ins → in four days

All terms undisclosed

UHY (Top 30; Summit Partners-backed) combined with Michigan CFO Associates of Clinton Township, Mich. — a fractional-CFO and advisory provider, founded 2006 — explicitly to expand its CAS practice. Eide Bailly (Top 20; Reverence Capital-backed since June) will add MUN of Roseville, Calif., effective Aug 17. Grassi (Top 60) added Hoffman Mulligan in New York City, effective Aug 1, expanding its tax practice.

→ Two things worth keeping. UHY's MICFO deal is its sixth of 2026 and buys a capability rather than a postcode — CAS is where firm-side AI is landing first, so watch whether capability buys start outnumbering geography buys. And Eide Bailly's MUN deal is its first tuck-in since Reverence took majority control in June — platform behaviour arriving roughly on schedule after a sponsor cheque. ✅ Also confirmed this week: UHY's backing is Summit Partners, since late 2024 — a standing verify flag on this tracker, now cleared.
Aug 3–6
2026

Five tuck-ins → in four days

All terms undisclosed

Frazier & Deeter (Top 50; General Atlantic-backed since April 2025) acquired Gray, Gray & Gray of Canton, Mass. — its first New England office, and its second deal of 2026. Sorren (Top 50; DFW Capital-backed) added Reno-based Persing Professional Group and Carefree Professional Practice Resources — its fifth and sixth of 2026. Mauldin & Jenkins (Top 60) merged in Schmoyer and Co. in Columbia, S.C., effective Aug 1. UHY (Top 30) added Anderson, Lower, Whitlow in Iowa. The Bonadio Group (Top 50) added Syracuse credit-risk specialist T. Gschwender & Associates.

→ A busy week with nothing to price. This is the ordinary texture of the roll-up between headline transactions: constant cadence, no disclosed figures, mostly in markets nobody writes about. ⚠️ A widely-repeated secondhand summary attributes "$198M revenue, 800 employees" to Gray, Gray & Gray. Those figures are not in the primary release and appear to be the acquirer's own scale, misattributed. Don't cite them.
Jul 29
2026

Grant Thornton Advisors → CBIZ

$5B enterprise value · all cash · $55.00/share (~54% premium to 30-day VWAP) (definitive agreement — go-shop to Aug 27, shareholder vote pending)

Grant Thornton Advisors LLC — the New Mountain Capital-backed non-attest entity — has agreed to acquire CBIZ, Inc. (NYSE: CBZ), with New Mountain investing incremental equity. The combined firm would be the fifth-largest US provider of professional services, tax and advisory services: more than $5B in domestic revenue, nearly $7.5B globally, 34,500+ professionals across 20+ countries. CBIZ's Benefits and Insurance Services segment will be separated into a stand-alone company backed by New Mountain. Close targeted for Q4 2026; CBZ would delist from the NYSE. The largest transaction of its kind in more than 25 years.

→ The two tracks became one transaction. The release says the deal "builds on" Grant Thornton Advisors' $1 billion, three-year AI investment — announced in September 2025, a year before this deal. New Mountain's managing director: it "allows Grant Thornton Advisors to rapidly bring its market-leading AI and technology platform deeper into the market." That inverts the model this tracker was built on. Not deal, then tech spend — tech spend, then a $5B deal to buy 9,500 more people to run it across. No EBITDA figure was disclosed, so no multiple can be derived from it.
Jul 28
2026

Capstone Partners → the sector's own numbers

Market data — not a deal

A record 194 transactions announced or completed in 2025 (+26% YoY), and 62 deals so far in 2026 (+14.8%). Financial acquirers — PE — now account for 54.8% of accounting-firm M&A, up from 38.9% in the prior-year period, with sponsor deal volume up 69.1%. Sector-focused PE fundraising is up 16.1% to $12.7B, and the median fund has grown 23.6% to $581M.

→ Strategic buyers no longer lead this sector; sponsors do, and their funds are getting bigger. Capstone's verdict on the technology question is the blunt one: "Technology transformation has become a strategic requirement in accounting services M&A, not a value enhancer." An AI-ready practice is now worth measurably more than an identical one that isn't.
Jul 17
2026

Baker Tilly → the public loan market

~$3B being marketed (terms not finalised — not yet priced or closed)

Deutsche Bank is preparing to market roughly $3 billion of debt for Hellman & Friedman-owned Baker Tilly Advisory Group, refinancing private credit — including the ~$1.5B package led by Blackstone, priced at 450bps over the benchmark, that helped fund the $7B Moss Adams merger (completed April 2025, creating the sixth-largest US firm). Investor meetings were expected the week of July 20. Baker Tilly, Deutsche Bank and H&F all declined to comment.

→ The first real public-market test of the thesis. Private credit prices a firm behind a closed door, among lenders with a relationship to protect. A broadly syndicated loan is priced by investors who have none. Every multiple in this tracker so far has been a private number — this one will be public. Watch the spread.
Jul 22
2026

Doeren Mayhew → McMillan, Whiteman & Sowers & Co.

Terms undisclosed · deals six and seven of 2026

The Audax-backed Top 50 firm added Punta Gorda, Fla.-based McMillan, Whiteman & Associates (effective June 30; 15+ audit and tax professionals) and Houston's Sowers & Co. (effective June 18). Leaders from both become principals. Doeren Mayhew has now integrated five firms into its Houston office in twelve months, taking local headcount to ~220 — a 58% increase since June 2025.

→ Seven deals in seven months, none with disclosed terms. This is what the tuck-in machine looks like once a sponsor is two years in: not headline deals, just relentless cadence in markets nobody writes about.
Jun 23
2026

Reverence Capital Partners → Eide Bailly

~$1.8B reported (value per reports; no EBITDA multiple disclosed) · majority

A majority stake in the Top-20 firm (Minneapolis; founded 1917; ~$800M revenue, 3,500+ staff, 50+ offices) goes to Reverence Capital Partners and co-investors. Current leadership stays; closing expected Q3. The ~$1.8B figure is from reporting, not the firm's own release.

→ ~$1.8B against ~$800M of revenue is roughly 2.25× revenue — a rich price for a firm a sponsor intends to make more efficient. The efficiency lever is technology, which is the whole reason Track B exists.
Jul 9
2026

Ascend → Kreischer Miller

Terms undisclosed

The Alpine Investors-backed platform adds the Horsham, Pa. firm — FY24 net revenue of $55.6M, 22 partners, ~220 staff serving Greater Philadelphia and the Lehigh Valley.

→ A firm at near-Top-100 scale, with strong organic growth, choosing a platform over independence. That's the tell: the model is no longer just absorbing firms that had to sell.
Jul 7
2026

Citrin Cooperman → LGA

Terms undisclosed · APS

The Blackstone-backed firm ($871M FY24 net revenue) takes Woburn, Mass.-based LGA and its ~150 professionals — the first Citrin tuck-in logged here since the $2B flip. Standard APS split: advisors entity takes non-attest assets, the CPA firm takes attest.

→ The flip thesis in action: buy at ~15×, then grow into the price. LGA's managing partner cited access to "investments in innovation and AI-enabled capabilities" — watch this space for the Track B follow-through.
Jul 7
2026

UHY → RBT CPAs

Terms undisclosed

The $385.5M Michigan-based IPA 100 firm combines with Newburgh, N.Y.'s RBT ($33.1M, 175+ professionals, 23 partners), taking the joint wealth-management practice to ~$1.5B in assets.

→ UHY has done 15+ acquisitions in recent years. The consolidation is not only a PE story — serial firm acquirers are running the same playbook.
Jul 7
2026

Unity Partners → Meaden & Moore

Terms undisclosed · APS

Cleveland firm founded in 1919; 250+ professionals across 15 offices in the US and UK. Its first institutional investor in 107 years. The deal includes an "Employee Purpose Plan" extending equity ownership to every employee — not only partners.

→ The standard objection to the roll-up wave is that partners cash out while everyone below them inherits the leverage. This is the first deal explicitly structured to answer it. Whether it becomes a template or a press release is the thing to watch.
Jul 1
2026

CohnReznick → Ham, Langston & Brezina

Terms undisclosed

Apax-backed CohnReznick absorbs the Houston firm (12 partners, 87 staff, founded 1995; energy, financial services, construction, non-profit, benefit plans), adding two Houston offices and one in Galveston. Its second acquisition of 2026, after Smith Schafer in January.

→ Textbook platform behavior: once the sponsor is in, the firm becomes an acquirer. Watch for auditor transitions among HL&B's public clients.
Jun
2026

KKR → Crowe

~$3B reported (verify) · APS

The headline deal. Crowe is the #12 US firm (~$1.39B revenue, 5,600+ staff) and was one of the last large holdouts. Majority stake; partners retain a minority. Split into Crowe LLP (attest) and Crowe Advisory LLC.

→ When a firm this size goes, the question stops being whether the model spreads and becomes who's left.
Jun
2026

TPG → Smith + Howard

Terms undisclosed

A ~$306B-AUM alternative asset manager enters accounting. Atlanta Top-100 firm (~$74M net revenue). Prior backer Broad Sky exits.

→ The mega-funds are now in. That changes the ceiling on what firms can fetch.
Jan
2026

Madison Dearborn → Nichols Cauley platform

Growth investment

MDP's first accounting deal. A three-way merger — CPA firm + Partners Risk Services (insurance) + JGH Consulting (transaction advisory) — into a single ~$49M-revenue platform, led by former Baker Tilly US CEO Alan Whitman.

→ Note the shape: not a firm, a platform. Accounting is the anchor, and adjacent high-margin services are bolted on.
Jan
2025

Blackstone → Citrin Cooperman

~$2B · ~15× EBITDA

The first PE-to-PE flip of a US accounting firm. New Mountain bought in at ~$500M (~11×) in 2021 and sold at ~$2B — roughly a 4× return in three and a half years. Revenue grew ~$350M → ~$850M over the hold. Blackstone kept its own stake below 50% for audit-independence reasons.

→ This is the deal that proves the exit. Every sponsor entering accounting is underwriting to this outcome — which is precisely why the pace has not slowed.
Mar
2024

New Mountain Capital → Grant Thornton US

60% of the non-attest business

Broke the size record weeks after Baker Tilly. The capital funded partner retirement obligations, returned capital to current partners, and built an M&A war chest.

→ The clearest statement of what the money is for: retire the old guard, pay the current one, and go shopping.
Feb
2024

Hellman & Friedman + Valeas → Baker Tilly

>$2B valuation

The deal that made PE-in-accounting mainstream — at the time, the largest US firm to take outside capital. Baker Tilly has since gone on the offensive itself (Moss Adams, Anchin).

→ Sponsor-backed firms don't stay bought. They become buyers.
2021
–2022

The first wave → EisnerAmper · Citrin Cooperman · Cherry Bekaert

TowerBrook · New Mountain · Parthenon

TowerBrook–EisnerAmper (2021) started it. New Mountain–Citrin Cooperman (2021, ~$500M/~11×) and Parthenon–Cherry Bekaert (2022) followed.

→ Three of the four fastest-growing US firms are now PE-owned. Whatever else the model does, it moves the growth needle.

Track B · What firms are buying

Technology adoption and investment by accounting firms: production deployments, announced spend, equity stakes in accounting-tech companies, vendor alliances — and the quiet retreats. Priority: the mid-market. Big Four AI spend is well covered everywhere; what a $50–500M-revenue firm actually deploys is not, and that's the firm most readers work in or compete against. Companies marked ATS100 appear on our AccountingTech Startup 100.

Aug 18
2026

Aprio launches Aprio Ventures

Nothing disclosed — no fund size, no committed capital, no portfolio names

Aprio — the 20th-largest US firm (3,400+ team members, clients in 50+ countries, Charlesbank Capital Partners-backed since July 2024) — launched Aprio Ventures, a formal corporate venture program sitting inside the firm's Corporate Development Team and led by Kyle Kling. It will invest across AI, automation and enterprise software, alongside established venture firms, pairing capital with technical support so Aprio professionals help shape how the tools work in practice. From the release: "Since 2019, Aprio has invested in nearly 20 companies." The program is explicitly framed as supporting the firm's five-year, $300 million technology and automation commitment.

→ The corporate venture arm has reached the mid-market — and it took two months. In June this tracker logged KPMG Ventures' equity portfolio (Incentify, Fieldguide, Ema, Wokelo, Rhino.AI) and wrote: watch for mid-market firms copying this. Aprio is the first. But the news isn't that Aprio started investing — it's that Aprio stopped doing it opportunistically. Nearly 20 investments over seven years, then a name, a leader, a website and a line on the org chart. That is the move from a partner's side interest to a firm capability with a budget. What it means for you: the firm across town may hold equity in the tool you're evaluating. Equity buys early access, roadmap influence and a second return if the thing works — and it costs neutrality. "Do you own any of it?" is now a fair question to ask a peer who recommends software. ⚠️ No portfolio companies were named, so there is nothing to cross-reference against the ATS100 — which is itself the finding. Undisclosed firm-held stakes are invisible validation, and we'll keep chasing the disclosures.
Aug 12
2026

Thrive Holdings raises $2B — and owns 50+ accounting firms

$2 billion at a $12 billion valuation · SoftBank, D1 Capital, Altimeter (performance figures below are self-reported)

Thrive Holdings — a spinout of Thrive Capital — raised $2 billion at a $12 billion valuation, its first outside capital after $1 billion from Thrive Capital's own backers. Its platform spans 70+ businesses across two pillars: Current, the accounting arm, with more than 50 firms and over 2,000 professionals; and Shield, its IT arm. OpenAI took an equity stake in the parent in December 2025 and embeds its own engineers in the portfolio — OpenAI's head of applied research holds a joint role at Thrive. The two co-built a tax-return processing agent on OpenAI's Codex, running inside Current as TaxAI.

→ This is a new bidder class, and it inverts the model this tracker was built on. A conventional sponsor buys a firm for its cash flow and hopes automation expands the margin later — the ~24-month lag we've now measured twice. Thrive buys firms as a deployment surface for automation that already exists, with the model vendor itself on the cap table. There is no lag, because the deal is the deployment. For a seller, the practical consequence is two structurally different bidders competing for the same practice, which is good for the price. ⚠️ Thrive says TaxAI "processed more than 7,000 tax returns at 98% accuracy," cutting prep time 30%+. Self-reported, no methodology published, and an earlier version of the claim specifies data-entry accuracy — a much narrower thing. Don't repeat "98% accuracy" unqualified.
Aug 11
2026

FloQast survey → ambition 85%, execution 10%

Readiness signal — not a deployment (vendor-sponsored research)

FloQast's State of Accounting AI 2026, across US and UK accounting and finance professionals: 85% of teams have made AI a strategic priority but only 10% use it extensively. 51% run close controls that are informal or inconsistently applied. Only 27% strongly agree current AI tools meet enterprise-grade governance and compliance standards. 92% of decision-makers expect AI investment to rise over two years — while just 17% say their teams are ready to put it to work. Reconciliations are named the top automation opportunity by 55%; only 5% report high automation there today.

→ The bottleneck is controls, not models — and that's now three independent readings. The CAQ found ~90% of audit partners rate their largest client's AI governance as developing or early stage (July). KPMG pulled its own agentic-AI report over fabricated citations (June). Now this. Different publishers, different samples, different motives, same conclusion: AI has to land on a documented, consistent control environment, and layering it over a weak one accelerates the errors instead of fixing them. Which is precisely why the dullest asset in the building is the one being bought in a roll-up. ⚠️ FloQast sells the controls-and-close layer this survey identifies as missing. Weigh it accordingly; sample size and methodology aren't published.
Aug 5
2026

Aprio co-builds audit agents with Fieldguide ATS100 #2

Co-development · terms undisclosed (agents under development — not yet a rollout)

Aprio — the 20th-largest US firm (FY25 net revenue $681.3M, 3,400+ team members, clients in 50+ countries) — is co-developing role-specific AI agents spanning the full audit lifecycle with Fieldguide: planning and risk assessment, fieldwork, review, financial statement preparation. Every workflow is built to comply with AICPA, PCAOB and GAAS standards, with human oversight at every stage, and each agent is shaped by Aprio's own audit methodology. The firm calls it "one expression of Aprio's $300 million, five-year investment in AI and automation."

→ The paper trail is what makes this worth logging. Charlesbank took Aprio's first institutional capital in August 2024; the $300M AI commitment and the TimeCredit acquisition landed July 2025; Fieldguide went across risk advisory in December 2025, into financial audit in May 2026, and to agents in August 2026. Sponsor cheque to agents in the audit file: two years, every step dated in a primary release. That's the second independent chain to land on ~24 months. ⚠️ Read the tense carefully — the Fieldguide platform is in production here; the agents are not yet. Co-development, not deployment.
Jul 23
2025

Aprio commits $300M — and buys TimeCredit outright

$300 million over five years · plus an acquisition (announced commitment — unaudited)

A five-year, $300 million commitment to AI and automation, announced alongside the outright acquisition of TimeCredit, an AI assistant company built by accountants. Aprio's release credits the Charlesbank Capital Partners investment of August 2024 — the firm's first institutional capital partner — as the enabler.

→ Treat the headline figure as you'd treat any multi-year press-release number: unaudited, with ordinary technology budget inside it. The acquisition is the harder signal. This is the "firm buys the capability outright" pattern this tracker has been hunting for — the first one logged. TimeCredit isn't on the ATS100 and, now firm-owned rather than independent, probably can't be. That's the point: some of these companies won't stay buyable.
Sep 4
2025

Grant Thornton Advisors commits $1B to AI

$1 billion over three years (announced commitment — unaudited, and partly software licensing)

AI tools and technology for the entire workforce of its multinational platform: Microsoft 365 Copilot rolled out to 13,500+ professionals across 60 offices, following a 400-person pilot. The firm's own release credits the New Mountain Capital investment as the source of the money, and notes CompliAI launched in May 2025.

→ Treat the headline number the way you'd treat any press-release figure — a multi-year global commitment with a lot of Microsoft licensing inside it. But it is now load-bearing. Logged here in August 2026 because the July 29 CBIZ release cites this spend as a rationale for a $5 billion acquisition. The AI came first; the scale to amortize it came second.
Jun 18
2026

KPMG takes equity in Incentify

Minority stake · size not disclosed · equity, not a pilot

KPMG LLP has taken a minority equity stake in Incentify, an AI-powered credits-and-incentives platform. Incentify's US incentives warehouse and AI discovery engine power ICON, embedded in KPMG Digital Gateway and covering programs across 120+ countries. ICON is described as already in use across KPMG's tax practice — production, not pilot. It builds on KPMG's exclusive global alliance with Anthropic.

→ The strongest form of Track B signal: a firm putting its own balance sheet behind a vendor. Incentify is not on the ATS100 — it's a candidate for Edition 2. The same release discloses prior KPMG Ventures stakes in Fieldguide ATS100 #2, Ema, Wokelo and Rhino.AI. Firm equity in accounting-tech is becoming a portfolio strategy, not a one-off — and a firm that owns a piece of its vendor has a very different incentive to make the deployment work.
Jul 14
2026

Grant Thornton Advisors runs CompliAI on Fieldguide ATS100 #2

Production deployment · terms undisclosed

Fieldguide's agentic platform is now the technology foundation behind CompliAI, Grant Thornton Advisors' AI-enabled controls and risk assessment tool. It supports delivery across the US Advisory practice, underpinning SOX compliance, fraud risk and other risk advisory engagements — an expansion of a multi-year collaboration, with plans to extend to further service lines and geographies. The counterparty is Grant Thornton Advisors LLC, the non-attest APS entity, not the CPA firm.

→ This closes the loop with a name and a date. New Mountain took 60% of Grant Thornton's non-attest business in 2024 with capital earmarked for technology. Two years later, that money is running agents on regulated, judgment-intensive work with liability attached. Track A to Track B, traced end to end — ~2 years, deal to deployment.
Jul 21
2026

KPMG becomes an OpenAI Elite Partner

Alliance · no dollar figure disclosed

The top tier of OpenAI's partner network. The alliance covers a "headless" enterprise model — AI as the system of engagement, applications as the system of record — plus modernization of KPMG's KRIS Connected state-and-local-government SaaS suite, participation in the OpenAI Daybreak cyber program, forward-deployed engineers with OpenAI certification, and KPMG as a client-zero deployment inside OpenAI's own operations.

→ The firm isn't only buying AI; it's being certified to implement and resell it. That turns the firm into a distribution channel, which is how vendors actually scale. Worth noting the calendar: this lands six weeks after KPMG retracted its own agentic-AI report over fabricated citations. Both things are true.
Jul 15
2026

The Center for Audit Quality counts the readiness gap

⚠️ Demand-side reality check · survey

A new CAQ survey of 589 audit partners finds nearly 90% rate their largest client's AI governance as "developing" or "early stage." AI governance topped the list of audit-committee challenges at 57% — ahead of cybersecurity — and the most-cited obstacle to controlling AI was "insufficient in-house AI expertise."

→ Track A pays a premium betting AI expands margins; the profession's own auditors say the governance to run that AI safely barely exists yet. The firm that can review and control an agent's output, not just deploy it, is the one that earns the multiple instead of paying for it.
Jul 1
2026

Sikich embeds Basis ATS100 #1

Production deployment · terms undisclosed

The Top 30 firm is putting Basis's autonomous AI agents into its client accounting services delivery — ERP integration, month-end close, transaction processing, reconciliation — across its global team.

→ The strongest class of signal there is: a real buyer, doing real diligence, carrying real liability, in production. Not a pilot, not a demo.
Jun 8
2026

Fieldguide ships Field Orchestrator ATS100 #2

Product launch · AICPA ENGAGE

A long-horizon agent that carries substantive testing end to end — population analysis, sample selection, evidence collection, extraction, testing, workpaper generation — pausing for practitioner review along the way. Shipped alongside Field Board (a kanban engagement view) and an Agent Review Experience.

→ The category shift, stated plainly: from single-task copilots to agents that sustain reasoning across a whole engagement. This is what "agentic audit" means once it stops being a keynote slide.
Jun 12
2026

KPMG pulls its own agentic-AI report

⚠️ Retreat — first one logged

The report, "Redefining excellence in the age of agentic AI," was withdrawn after research group GPTZero found only 5 of its 45 citations pointed to real, intact sources — the rest fabricated by the AI that helped write it. UBS, the NHS, Swiss Federal Railways, and Transport for London all disputed claims about their AI usage.

→ Everyone reports the launches; this column exists for the failures. The firms selling AI transformation are still learning to review its output — the strongest argument yet that the review layer is the job.
2023
–2026

The Big Four buy platforms, not tools

PwC ~$1B · KPMG ~$2B · EY ~$1.4B (all reported; verify)

PwC partnered with OpenAI and deployed ChatGPT Enterprise firm-wide at reported six-figure license counts, becoming a reseller to clients in the process. KPMG committed to a multi-year Microsoft AI alliance. EY built EY.ai and its EYQ internal model. Deloitte moved to agents with its Zora platform and an Nvidia partnership.

→ Treat these headline figures as press releases, not financial statements — they are unaudited, and some is re-labeled IT budget. The deployment count is the real number. The more interesting question, and the one this tracker exists to answer, is what everyone below the Big Four actually adopts.
2024
–2026

Grant Thornton (New Mountain) earmarks capital for technology

Part of the sponsor thesis (verify)

PE capital explicitly directed at technology alongside the M&A war chest.

→ Track A meeting Track B in a single sentence: the sponsor's thesis is the tech spend.
Watch
list

What we're hunting next

Equity, not just licenses. When a firm takes a stake in a vendor — as Citrin Cooperman did with Tellen — it is a far stronger signal than a pilot. Firms buying tech companies outright. The "buy the capability" route. And the retreats: abandoned pilots, quiet write-offs, tools that got dropped.

→ Everyone reports the launches. Almost nobody reports the failures — and the failures are more useful to anyone deciding what to buy. If you've watched a rollout die inside a firm, we want to hear about it.

What the data says so far

Updated as the database grows. These are working conclusions, and we'll say so plainly if the evidence turns against them.

🕵️ Know something before it's announced?

Practitioners always do. Deals leak, pilots quietly die, and rollouts get walked back long before a press release admits it. If you've seen one from the inside — a deal in diligence, a tool that flopped, a partner vote that failed — tell us. We protect sources, and we verify everything before we print it.

Send a tip

The other half of the story: the ATS100

Track B tells you which technology firms are buying. The AccountingTech Startup 100 tells you who's building it — 100 companies across 13 categories, with funding signals. The two lists check each other: when a firm puts money behind a company on the list, the list is doing its job.

See the ATS100 →

Get the tracker every Monday

The week's deals, the week's deployments, and what they mean for your firm. Free, and about two minutes to read.