KPMG is testing an artificial intelligence-powered simulation platform designed to help junior tax employees develop skills that traditionally took years of repetitive client work to acquire, underscoring how large professional-services firms are rethinking workforce training as AI rapidly automates entry-level tasks.
The system, called TaxSIM, is being developed with Centaurian AI and is expected to roll out later this year to KPMG’s roughly 10,000 tax professionals in the United States.
The initiative reflects a growing concern spreading across white-collar industries. As AI absorbs more routine work, younger employees may lose the repetition and hands-on exposure that historically formed the foundation of professional expertise.
For accounting, consulting, and legal firms, the issue is becoming increasingly urgent because many traditional apprenticeship models depend heavily on junior workers repeatedly performing labor-intensive tasks before progressing into advisory and strategic roles.
Brad Brown, KPMG’s chief digital officer for tax, said the company began reassessing its training structure as generative AI systems started automating large portions of tax preparation work.
“You’re not going to get as many repetitions of doing that task as you would have in the past,” Brown said. “So we needed something to fill that void.”
Before the rise of AI-driven automation, early-career tax professionals often spent several years preparing client returns manually, gradually building technical judgment and pattern recognition through repetition.
Brown said junior employees at KPMG traditionally spent roughly four years working through tax returns before transitioning into higher-level advisory functions involving regulatory strategy, business structuring, and tax planning.
That process is now changing rapidly.
As AI systems increasingly handle document review, compliance checks, and return preparation, firms are confronting the possibility that younger workers could advance into senior roles without developing the practical instincts that come from years of repetitive work.
AI Simulation Becomes the New Corporate Training Ground
KPMG’s response is to compress years of learning into simulated high-volume experiences designed to mimic real-world complexity. The TaxSIM platform allows employees to move through large numbers of tax scenarios at an accelerated speed, generating different outcomes depending on user decisions and analytical reasoning.
Brown said the goal is to help workers build professional judgment “much more rapidly” than under traditional training structures.
“It just gives us incredible acceleration to create those skills,” he said.
The platform mirrors developments already underway in other industries where simulation-based learning has replaced or supplemented real-world repetition.
Kes Sampanthar, cofounder and CEO of Centaurian AI, which developed the tool with KPMG, compared the concept to advanced motorsport simulators used by professional drivers.
“It’s like the top athlete who gets better and better if they can get the right feedback,” Sampanthar said, referencing the racing simulation game Gran Turismo.
Instead of waiting years to encounter a broad range of client situations organically, employees can cycle through numerous simulated tax and business scenarios in compressed timeframes. The platform is also designed to expose workers to macroeconomic and geopolitical variables that younger professionals may otherwise rarely analyze directly.
Sampanthar said users can test how factors such as tariffs, changing regulations, or economic shifts ripple through corporations, industries, and national tax systems.
“These are things you normally don’t get feedback on,” he said.
The broader implication is that AI is not only automating work itself but also changing how expertise is formed inside corporations. Historically, many professions relied on repetition to build instinctive decision-making. Junior bankers built financial models repeatedly. Lawyers reviewed contracts for years. Consultants assembled endless presentations and spreadsheets. Tax analysts prepared thousands of returns.
Now, AI systems increasingly perform many of those tasks faster and at lower cost. That raises a difficult structural question confronting employers globally: how do workers develop judgment when machines handle most of the repetition?
Firms Race to Preserve Human Judgment in an AI Economy
KPMG’s experiment highlights a growing consensus among large employers that future workforce value may depend less on manual execution and more on analytical reasoning, oversight, and decision-making.
Brown said younger workers are already expressing anxiety about how career development changes in an AI-heavy environment. But he argued that the pressure to adapt extends across all levels of the workforce, not just junior staff.
“The need to continually learn cuts across every rung of the ladder,” Brown said.
The simulation platform intentionally requires users to reason through problems before turning to AI-generated assistance, an effort aimed at preventing overreliance on automation.
“Learning happens when things are hard, not when things are easy,” Sampanthar said.
The emphasis reflects broader concerns emerging across corporate America and the technology sector that workers may become dependent on AI tools without fully understanding the underlying processes. Several large firms are now experimenting with hybrid training approaches that combine limited hands-on work with AI-assisted learning and simulation environments.
Brown said KPMG does not expect simulations to fully replace foundational experience. He recalled a recent conversation with a first-year tax analyst who still wanted to manually build a valuation model before relying on simulation tools.
According to Brown, the employee’s view was: “It’s OK to do one or two, but I don’t need to do four years of these to get to that level of skill.”
That mindset increasingly captures the direction many professional-services firms are moving toward: preserving enough manual exposure to build a foundational understanding while using AI to dramatically accelerate the path toward higher-value analytical work.
The shift could ultimately reshape not only corporate training but also the structure of white-collar employment itself.
For decades, industries such as accounting, consulting, and finance relied on large pools of junior workers performing repetitive tasks as part of a long apprenticeship pipeline.
AI is beginning to dismantle that model.
In its place, firms are attempting to build a new system where workers spend less time on execution and more time supervising AI systems, interpreting outcomes, and advising clients on increasingly complex business decisions. KPMG’s TaxSIM initiative may offer one of the clearest early signals of how that transition is beginning to take shape inside major corporations.
