The impact of AI on the accountancy profession?

Author’s Note

One of the abilities I am keen to explore with AI is how it can help me to write. This article was co-authored with ChatGpt5.2 and reviewed by Claude Sonnet 4.6 - I will write an article on my experiences when I have some more experience of working with it.

Executive Summary

AI will not transform accounting firms as they are currently structured. It will make them more efficient and compress margins on routine work, but it will not, by itself, change the underlying partnership‑and‑billable‑hours model.

Every major technology wave before AI — computers, desktop accounting software, ERP, practice management, cloud platforms, RPA — promised transformation and delivered incremental improvement instead. The limiting factor was structure, not technology.

AI’s real disruptive potential sits in new structures: private‑equity‑backed platforms and AI‑native providers that industrialise bookkeeping and basic compliance. These are the “production lines” that enable economies of scale and, ultimately, consolidation.

Many commentators assume end‑clients are organised and tech‑literate. They are not. Much of the manual pain is still in cleaning up messy bookkeeping and close processes — exactly the area firms most want to automate and the area new platforms are targeting first.

For both small and mid‑sized clients, AI‑enabled platforms only need to offer visibly better economics on the “button‑press” work and a story that fits the client’s perception of value for switching to become attractive. Traditional firms can respond — but only if they are willing to change ownership, incentives and the status of bookkeeping.

Article

Why this isn’t another “AI will change everything” article

When I talk to partners of accounting firms about AI, one of things they often say is “we’ve been here before and nothing substantially changed”. This article starts from that perspective and asks a harder question: even if AI is different, will our structure let it change very much?

Over the last forty years, UK accounting firms have been told that mainframes, PCs, accounting software, ERP, practice management, cloud and RPA would all “transform the profession”. Each wave brought a new set of demos, a new set of consultants and a lot of confident predictions. Each wave also delivered something real. Audit files moved off paper. Ledgers moved off manual books. Cloud platforms reduced the need to chase clients for backups. RPA scripts saved time on particularly mind‑numbing tasks (Hubifi, 2025).

None of those waves fundamentally changed what most firms are: partnerships selling time and judgement, with pyramids of staff feeding work upwards. That isn’t because the technology was weak. It’s because the model was very good at absorbing technical change without changing itself.

The risk with AI is not that it will do nothing. The risk is that we repeat the same pattern: a lot of activity, some real gains, and no structural shift in how firms create, deliver and capture value.

What the last four decades of “transformation” actually did

The first wave was the computer itself. From the 1960s into the 1980s, large organisations moved ledgers and payroll onto mainframes and minicomputers, often via in‑house “EDP” departments. That was a big deal for them; for most practitioners it was invisible.

In the 1980s and 1990s, PCs arrived on every desk. Desktop accounting packages and spreadsheets made it much easier for SMEs to keep books and for accountants to swap files instead of paper. The work was faster and less error‑prone. The outputs were the same.

Larger clients installed ERP systems that integrated accounting with operations, giving richer data but more complex reconciliations. Firms installed practice management systems that tracked clients, WIP, time and billing. Those helped us run firms more professionally. They did not change the basic deal: clients pay for work; partners distribute profits.

Cloud platforms like Xero, QuickBooks Online and Sage Business Cloud were a more significant shock, especially in the SME space. Real‑time feeds, shared access and app ecosystems genuinely changed how outsourcing and bookkeeping worked. Firms with more clients on cloud tended to have higher revenue per client, which suggests that standardising on modern tools does matter (Xero, 2024). But again, the model flexed. We did the same jobs, just at slightly different frequencies and price points.

RPA came later. It automated some repetitive work on top of whatever systems we already had. Helpful, yes. Transformative, no. In many cases it simply papered over the cracks between legacy systems rather than prompting anyone to rethink the underlying architecture (Trintech, 2023).

If you step back, the pattern is stark: new technology, real benefit, no structural change.

Why those waves didn’t change the model

So why didn’t any of that “transform the profession”?

You can tell the story in four constraints.

Conservatism and liability. Our regulatory and liability framework rewards caution. Any new technology that looks like a black box triggers a reasonable instinct to go slowly. One bad failure can wipe out years of gains (ICAEW, 2025a).

Client expectations. Historically, clients valued relationships and judgement more than tooling. They didn’t reward firms for putting more technology behind the scenes; they rewarded them for being available and careful. That made it rational to treat technology as plumbing, not as something that changed the proposition (Susskind and Susskind, 2015).

Fragmented data and systems. Clients and firms both run multiple systems and spreadsheets. Getting clean, consistent data is hard. Many “intelligent” tools ran out of road in the real world. Empirical research using the TOE (Technology-Organisation-Environment) framework confirms that AI readiness, regulatory environment and competitive context all materially moderate adoption rates in professional services firms, with smaller firms disproportionately disadvantaged (Sjödin et al., 2024).

Incentives. The billable hour and short‑term partner profit share do not reward cannibalising today’s revenue for the sake of tomorrow’s model. Technology teams have been cost centres; optimisation was rewarded, reinvention wasn’t.

Those four constraints are still there. AI does not magically remove any of them. Research using Big Four audit firms as the study population confirms that professional services organisations typically respond to digital disruption through “creative accumulation” — absorbing new tools into existing structures rather than restructuring around them (Sele and Grand, 2023).

AI inside today’s firms: powerful, but contained

AI has two obvious advantages over previous waves:

• It’s easier to use. You can talk to it. You don’t need to learn a new reporting tool or scripting language (AccountingWEB, 2026; ICAEW, 2025b).

• It’s more capable. It can work with unstructured text, generate drafts, summarise documents and help developers write and test code faster (MIT Sloan, 2025).

Those advantages matter. They mean you can:

• Give every junior an assistant that drafts memos, working paper notes and emails.

• Use copilots inside audit, tax and practice tools to speed up review work (AccountingWEB, 2026).

• Help your own developers and vendors modernise legacy systems faster than before.

However, regulators, professional bodies and insurers still insist that AI must augment, not replace, professional judgement. Firms still carry full responsibility for outputs. Insurers still worry about model validation, data leakage and control failures (ICAEW, 2025b; CFOTech, 2025).

Inside the current model, that combination leads to three outcomes:

• AI is deployed cautiously, mainly in non‑signature work and vendor tools.

• Productivity improves.

• Margins on routine work come under pressure as clients push back on paying yesterday’s prices for what they increasingly see as commoditised.

That is not nothing. For many firms it will be the difference between flat profits and declining profits over the next decade. But it is not transformation.

So what could actually drive transformation?

There are two groups of people outside accounting firms who have the power to drive industry‑level change:

• Private equity investors.

• Clients.

Private equity and the economics of scale

Private equity investment in accounting and related services has risen sharply. A significant proportion of firms say they are open to PE investment; several mid‑tier practices have already taken capital and started to build groups (ICAEW, 2025a; Accountancy Age, 2025).

PE‑backed groups tend to do three things:

• Centralise operations and technology across multiple local brands.

• Invest heavily in automation to expand margins and capacity.

• Use acquisitions to feed volume into that central “factory”.

We have seen this dynamic across capital-intensive industries since the industrial revolution. In automotive manufacturing, the moving assembly line at Ford’s Highland Park plant — introduced on 7 October 1913 — reduced Model T assembly time from 12.5 hours per car to just 93 minutes (Hounshell, 1984, pp. 217–261). Unit prices fell from approximately $850 at launch to around $260 by the mid‑1920s. Crucially, as Hounshell demonstrates, the assembly line’s productivity gains were inseparable from radical product standardisation: the Model T was deliberately held constant to allow the process to be optimised around it (Hounshell, 1984, p. 220). Those economics drove consolidation across the industry.

The broader economic logic was captured by Chandler (1990) in his analysis of industrial capitalism: firms that made the “three-pronged investment” in production capacity, distribution networks and management capability at sufficient scale achieved first-mover advantages that later entrants found almost impossible to overcome (Chandler, 1990, pp. 14–46). Scale and scope economies, once captured, became self-reinforcing barriers to entry.

Private equity has a reputation for aggressively maximising return on investments, and there is no reason to believe that this will be any different in accounting (Thomson Reuters, 2026). This gives them a choice: try to maintain prices while increasing volume and reducing costs, or go for market share with economies of scale compensating for reduced pricing. We are already seeing PE‑backed groups building offshore centres or buying outsourcers; inevitably they will try to standardise platforms and automate to further increase their returns. The trade‑off is simple: where partnerships hesitate to industrialise work that feels “professional”, PE owners have no such emotional attachment.

There is, however, an important caveat: accounting is not the Model T. Ford’s production line worked because the product was standardised. Accounting and compliance work varies by client sector, size, complexity and jurisdiction. PE-backed roll-ups that achieve genuine industrialisation are therefore different from those that are simply collections of acquired partnerships under a common brand — and that distinction matters enormously for whether the economics Chandler describes actually materialise.

Traditional partnerships can try to copy parts of this. PE‑backed platforms that achieve genuine standardisation are structurally designed for it.

Clients and the perception of value

Clients, meanwhile, increasingly assume you’re using AI whether you are or not. Many will quietly expect faster turnaround and lower prices on routine work because, in their heads, you’re “just pressing a button” (Wolters Kluwer, 2025).

Micro and small businesses are under increasing pressure to cut costs to stay afloat. It is not unreasonable for them to seek a lower‑cost accountant. New rules like MTD may accelerate that by forcing them to buy software and engage more digitally anyway (Simply Business, 2025; HMRC, 2025).

For a mid‑sized business, the calculation is more complex, but not fundamentally different:

• They already pay substantial monthly or annual fees for full‑service accounting (Ask Accountants UK, 2025).

• Their own bookkeeping and close processes are often messy (Financier Worldwide, 2024).

• They can see that AI and automation should be able to do more of the heavy lifting.

Will they increasingly expect their accountant to support full automation of their finance processes? It would be surprising if they did not. The question we have to ask is whether new entrants to the industry, perhaps with private equity backing, can exploit this perception.

Mid‑sized end‑clients — say £5m–£250m turnover — are not going to stampede to the cheapest online accountant. They care about relationships, sector knowledge and risk.

Even so, three things are likely to move them:

Price and package differences on the basics. If an accounting firm can bundle monthly bookkeeping, management accounts and compliance at a visibly better effective rate than the current mix of fees, a finance director can justify switching as a rational cost decision, not a risky experiment (Ask Accountants UK, 2025).

Reliability and responsiveness. Many mid‑sized businesses have lived through late filings, unreturned calls or partner churn. A provider that offers real‑time status, fewer surprises and clear SLAs will look attractive.

Alignment with their own AI and data ambitions. As mid‑sized businesses put AI into their own finance functions — for invoice capture, forecasting or dashboards — they will expect their accountants to at least keep pace. A firm that still runs the engagement off emailed spreadsheets looks increasingly misaligned (Financier Worldwide, 2024).

None of this means that every mid‑sized client will move. Many won’t. But it does mean there is a market opportunity through which a competitor can entice them to move — and that route runs through AI‑enabled bookkeeping and monthly close processes, not through flashy advisory work.

Bookkeeping, the Cinderella of accountancy

If you look critically at finance processes, there are numerous inefficiencies in compliance work due to issues in bookkeeping. Garbage in, garbage out. If you ask accounting firm partners which part of their work they most want to automate, bookkeeping and monthly/annual close processes are usually near the top. They are labour‑intensive, price‑sensitive and hard to leverage.

The automation industry — both cloud ledger products and AI‑centric firms — plays to this. However, there is a quiet assumption behind a lot of commentary: that most business owners are reasonably organised and reasonably comfortable with software. They are not. Records arrive late, in multiple formats, with personal and business items mixed together. Many mid‑sized finance teams are not much better: they juggle ERPs, sector apps and spreadsheets, with a lot of manual adjustment at month‑end (Quality Brains, 2025; Sage, 2025; Financier Worldwide, 2024).

This is exactly where a new generation of providers is focusing. A wave of AI‑enabled bookkeeping and compliance platforms has emerged that combines heavy automation with centralised human teams to deliver fixed‑fee services. They offer all‑in packages that assume clients will not show up with immaculate records or deep software skills (Quality Brains, 2025; Wolters Kluwer, 2025). They start where the pain is worst and the prestige is lowest: the “Cinderella” bookkeeping work that keeps everything else running. They build workflows and AI tooling to absorb that mess, and they build recurring revenue on top of the very work many firms treat as a low‑status training ground.

If one of these new entrants can say, credibly, “we’ve industrialised the button‑press work you already think is commoditised, and we’ll give you predictable pricing and better visibility”, the barriers to switching drop. And if you believe — as many clients now do — that much of the work is really “system work” rather than expert work, it becomes quite hard to argue that they should pay significantly more for the same outputs from a traditional firm.

Where does this leave traditional firms?

If you put all of this together, three things follow.

• AI will improve your current model. It will not fix it. Inside the existing partnership and incentive structure, AI will make you more efficient and will help you cope with MTD, regulatory change and staff shortages. It will also accelerate client expectations about speed and price on routine work. It will not, by itself, change the basic economics of your firm (Simply Business, 2025; HMRC, 2025).

• The real disruption, if it comes, will be structural. PE‑backed platforms and AI‑native entrants are building the equivalent of production lines for bookkeeping and compliance. Those that achieve genuine standardisation — not merely a rebadged collection of acquired partnerships — are designed to exploit the same economies of scale that Chandler (1990) identifies as historically decisive in capital-intensive industries (PKF Littlejohn, 2025; Thomson Reuters, 2026; ICAEW, 2025a).

• The danger zone is the squeezed middle. At one end, specialist firms with deep niches and high‑touch advisory models can continue to do well, using AI to support rather than define their value. At the other end, platforms can compete on cost and scale. The firms most at risk are those that are neither: generalist, mid‑sized, and unwilling to change structure. As Susskind and Susskind (2015, p. ix) observe, it is precisely the middle layer of professional work — competent but not exceptional, routine but not yet automated — that is most exposed to decomposition by technology-enabled new entrants.

The questions this really forces

For partners and IT directors, the interesting questions are not about which AI copilot to buy. They are about structure.

• Are we content to remain a traditional professional services firm, using AI as an efficiency tool and accepting that margins on routine work will come under pressure?

• Do we want to move towards a platform model ourselves — with centralised, industrialised bookkeeping and compliance — and if so, are we prepared to change ownership, governance and incentives to make that possible?

• How seriously are we prepared to take the “Cinderella” layer? Are we willing to put capital and senior attention into industrialising bookkeeping and close processes, rather than treating them as low‑status work for juniors?

• Finally, what kind of clients do we want in five to ten years’ time — and what kind of firm will they expect to be working with?

AI will not transform accounting firms as they are currently structured. It may, however, make it much easier for new structures to emerge that do not look like the firms we know today — and that may trigger an existential crisis. Whether we avert that is a strategic choice, not a technological inevitability.

References

Accountancy Age (2025) Half of UK accountancy firms open to private equity. Available at: https://accountancyage.com/2025/07/22/half-of-uk-accountancy-firms-open-to-private-equity/ (Accessed: 30 January 2026).

AccountingWEB (2026) 4 smart ways accountants can use AI in 2026. Available at: https://www.accountingweb.co.uk/tech/tech-pulse/4-smart-ways-accountants-can-use-ai-in-2026 (Accessed: 30 January 2026).

Ask Accountants UK (2025) Small business accountant cost in 2025: what you should really expect to pay. Available at: https://askaccountants.co.uk/small-business-accountant-cost/ (Accessed: 30 January 2026).

CFOTech (2025) AI to redefine finance as cost and cyber risks surge in 2026. Available at: https://cfotech.co.uk/story/ai-to-redefine-finance-as-cost-cyber-risks-surge-in-2026 (Accessed: 30 January 2026).

Chandler, A.D. (1990) Scale and scope: the dynamics of industrial capitalism. Cambridge, MA: Belknap Press of Harvard University Press.

Financier Worldwide (2024) AI reshaping corporate finance. Available at: https://www.financierworldwide.com/ai-reshaping-corporate-finance (Accessed: 30 January 2026).

HMRC (2025) HMRC’s transformation roadmap. Available at: https://www.gov.uk/government/publications/hmrc-transformation-roadmap/hmrcs-transformation-roadmap (Accessed: 30 January 2026).

Hounshell, D.A. (1984) From the American system to mass production, 1800–1932: the development of manufacturing technology in the United States. Baltimore: Johns Hopkins University Press.

Hubifi (2025) RPA in accounting: automating repetitive financial tasks. Available at: https://www.hubifi.com/blog/rpa-in-accounting (Accessed: 30 January 2026).

ICAEW (2025a) Why the practice finance landscape is changing. Available at: https://www.icaew.com/insights/viewpoints-on-the-news/2025/jan-2025/why-the-practice-finance-landscape-is-changing (Accessed: 30 January 2026).

ICAEW (2025b) Will 2026 be AI’s year? AI adoption trends. Available at: https://www.icaew.com/insights/viewpoints-on-the-news/2026/jan-2026/will-2026-be-ais-year-ai-adoption-trends (Accessed: 30 January 2026).

MIT Sloan Management Review (2025) How generative AI can make accountants more productive. Available at: https://mitsloan.mit.edu/ideas-made-to-matter/how-generative-ai-can-make-accountants-more-productive (Accessed: 30 January 2026).

PKF Littlejohn (2025) UK private equity market 2025: performance and outlook. Available at: https://www.pkf-l.com/insights/uk-private-equity-market-2025-trends-outlook/ (Accessed: 30 January 2026).

Quality Brains (2025) How AI is changing bookkeeping for small businesses. Available at: https://qualitybrains.co.uk/ai-changing-bookkeeping/ (Accessed: 30 January 2026).

Sage (2025) How bookkeepers are the human API powering UK small businesses. Available at: https://www.sage.com/en-gb/blog/bookkeepers-human-api-uk-small-businesses/ (Accessed: 30 January 2026).

Sele, K. and Grand, S. (2023) ‘Defusing digital disruption through creative accumulation: technology-induced innovation in professional service firms’, Journal of Management Studies, 60(7), pp. 1801–1835. Available at: https://doi.org/10.1111/joms.12972 (Accessed: 27 February 2026).

Simply Business (2025) 72% of businesses expect MTD to increase their costs. Available at: https://www.simplybusiness.co.uk/knowledge/self-assessment/how-much-will-making-tax-digital-cost/ (Accessed: 30 January 2026).

Sjödin, D., Parida, V., Palmié, M. and Wincent, J. (2024) ‘Artificial intelligence adoption in a professional service industry: a multiple case study’, Technological Forecasting and Social Change, 200, art. 123251. Available at: https://doi.org/10.1016/j.techfore.2024.123251 (Accessed: 27 February 2026).

Susskind, R. and Susskind, D. (2015) The future of the professions: how technology will transform the work of human experts. Oxford: Oxford University Press.

Thomson Reuters (2026) How PE can accelerate technology and enable accounting firm growth. Available at: https://www.thomsonreuters.com/en-us/posts/tax-and-accounting/pe-enable-tech-growth/ (Accessed: 30 January 2026).

Trintech (2023) The inevitability of RPA in finance and accounting: ubiquity of cloud computing. Available at: https://www.trintech.com/blog/the-inevitability-of-rpa-in-finance-and-accounting-ubiquity-of-cloud-computing/ (Accessed: 30 January 2026).

Wolters Kluwer (2025) How are accountants in the UK using artificial intelligence? Available at: https://www.wolterskluwer.com/en-gb/expert-insights/how-is-ai-being-used-by-accountants-in-the-uk (Accessed: 30 January 2026).

Xero (2024) Key insights from Xero’s UK accounting and industry performance report. Available at: https://blog.xero.com/data-insights/xero-uk-2021-industry-performance-report/ (Accessed: 30 January 2026).

Bibliography

Brenner, R. and Glick, M. (1991) ‘The regulation approach: theory and history’, New Left Review, 188, pp. 45–119.

Furneaux, B. and Wade, M. (2011) ‘An exploration of organizational motivations for enterprise system retirement’, MIS Quarterly, 35(3), pp. 851–876. Available at: https://doi.org/10.2307/23042607 (Accessed: 27 February 2026).

Viaante (2025) AI and automation in accounting: top UK trends to watch in 2026. Available at: https://www.viaante.com/blog/ai-automation-accounting-uk-trends-2026 (Accessed: 30 January 2026).