KPMG’s AI push is shrinking the accounting ladder from both ends
For years, professional services firms followed a familiar model: hire many graduates, assign them repetitive tasks, bill clients for their work, and gradually promote a few into leadership. This approach relied on size, hierarchy, and time.
Artificial intelligence is beginning to disrupt this model. Entry-level tasks are now becoming automated.
KPMG’s recent actions highlight this shift. The firm reportedly cut about 10% of its U.S. audit partners while also speeding up its use of AI in audits. These decisions are linked, showing the profession is moving from labor-heavy growth to a model focused on specialized expertise.
The Event
Routine testing has long been the foundation of early accounting careers. Associates learned by reviewing transactions, checking documents, tracing balances, and handling many small inconsistencies. The repetitive work helped build their judgment.
KPMG’s stance shows the firm no longer sees this work as an efficient use of people’s time.
At the same time, the partner cuts reveal another side of the transition. The reductions were not described as broad-based retrenchment; instead, they frame the pressure as falling on partners whose economics depended heavily on leverage rather than client origination or differentiated expertise.
This difference matters. AI doesn’t affect all parts of a firm equally. It changes which skills and roles are most valuable.
The Real Driver
The main reason is efficiency. Firms automate repetitive tasks because software does them faster and cheaper. But that’s only part of the story.
A bigger issue is that AI undermines the traditional economic model of accounting firms.
For years, firms made profits by having layers of staff. Junior employees did routine tasks, managers checked their work, and partners earned the difference between labor costs and billing rates. This system relied on lots of people at the bottom.
AI is aimed directly at that layer.
When software takes over routine testing, document review, and other structured tasks, firms need fewer people for those jobs. This changes how firms hire, promote, and even structure partnerships.
As the lower-level work becomes easier to automate, the value of top-level roles increases.
Clients still want experienced professionals to make tough decisions. Boards still value judgment when facts are unclear or sensitive. Reputation is still important when problems arise. AI can speed up analysis, but people are still responsible for the outcomes.
The term “K-shaped” describes a situation in which one part of the profession is climbing rapidly while the other is declining, illustrating how different groups may experience very different economic outcomes.
The top tier benefits because technology lets successful professionals do more. A trusted partner can oversee more work, respond faster, and provide advice more efficiently with AI tools to support them. The lower tier faces a different result. Standardized compliance work becomes cheaper and easier to commoditize. The same tools that help the top reduce the need for workers at the bottom. Managers are no longer supervising large groups performing manual procedures—their role shifts toward coordinating, overseeing, and interpreting outputs generated by automated systems.
This is a redesign of how firms create and distribute economic value.
The Pattern
Firms that have fully adopted AI are pulling ahead of those still treating it as a trial. This divide is about incentives as much as it is about technical skills. Well-funded organizations, or those backed by private equity, can move faster because their leaders have more incentive to restructure early. These firms are already redesigning workflows, building their own tools, and changing pay structures to attract the partners they think will be important in the future.
In contrast, older partnerships face pushback from within. Senior partners close to retirement often have little reason to change a system that still works for them. This slows down investment and delays change inside professions. The article also notes another common pattern in professional services: a gap between what firms say publicly and what they actually do internally. Internally, actual usage can look far less consistent. The gap becomes especially important when leadership itself uses the tools less frequently than the staff does. In that situation, governance problems emerge quickly because executives are making strategic decisions about systems they do not fully understand operationally.
This pattern shows up often during times of technological change. Institutions usually don’t struggle because technology fails, but because their incentives, authority, and habits change more slowly than the technology does.
Implications
The immediate consequence is that the traditional career ladder inside accounting firms becomes narrower and eventually nonexistent.
Entry-level work used to serve two roles: it produced billable work and trained future professionals. If AI reduces the need for this work, firms may also lose the process that builds experienced judgment over time.
This leads to a long-term succession problem.
Audit judgment is developed through repeated exposure to low-risk decisions before professionals handle bigger ones. If firms remove much of this repetitive work, they may end up with too few people who have deep practical experience.
The industry’s pricing structure is also changing.
Routine work is now faster, cheaper, and harder to tell apart. Clients who buy standard compliance services will likely push harder on fees because technology makes labor less visible.
At the top end, pricing power may strengthen instead of weaken. Professionals with strong industry credibility, specialized expertise, and durable client trust become harder to replace because those qualities do not scale as easily as automation can.
In the new model, the main products are: deep industry experience, judgment, trust, and personal reputation.
Forward View
The accounting firms will not disappear. There is still a need for assurance, oversight, and professional judgment. What’s changing is how firms organize their people around these roles. It is a thinner professional structure supported by heavier automation underneath it. More output will come from fewer people, but the remaining people at the top will carry greater economic importance.
This also explains why private equity investors now judge firms by how seriously management has taken AI adoption and operational redesign. The question isn’t whether automation affects professional services, but whether leaders acted soon enough to reshape the firm before the economics changed.
Accounting firms that adapt may still look the same from the outside, but inside, they will run on very different ideas about labor, expertise, and value.