Are We Automating Professional Services Into a Knowledge Crisis?

The Expertise Crisis Hidden Inside AI Adoption 

As AI adoption accelerates in professional services, firms risk losing judgement and expertise. Why cognition, not automation, should dominate AI strategies in 2026. 

In 2026, professional services firms will face an unexpected reckoning. AI will be well embedded into law, consulting, finance, accountancy, and government-adjacent work. Productivity will be up. Turnaround times will be down. The numbers confirm this shift: Thompson Reuters found that firms’ use of generative AI doubled in 2025, and that 95% of professionals believe AI will soon be central to their workflows.  

As AI takes hold, organisations will feel the effects of something vital to their success slipping away. That “something” is expertise. 

Hyper-focus around the potential for AI to replace humans means we are losing sight of a more pressing, near-term problem: the risk that AI removes the experiences through which professionals learn how to think. 

Most AI implementations across professional services have been designed around speed, efficiency, and cost reduction. Pattern recognition tasks are automated. Information retrieval is instant. Outputs are cleaner and faster. But this approach creates a dangerous blind spot: if early- and mid-career professionals are no longer exposed to the cognitive work behind critical thinking and decision-making, where will the senior professionals of tomorrow come from? 

The defining challenge of AI in professional services in 2026 is not improving technical capability. It is whether firms can adopt AI without hollowing out the judgement, intuition, and strategic reasoning that make professional advice valuable in the first place. 

In both cases, the solution is not to slow AI adoption. It’s to rethink what AI can and should accomplish in professions where expertise is the currency that drives firms’ financial success.  

What Professional Expertise Really Is — and Why AI Struggles to Capture It 

Expertise develops as much through experience as from formal instruction. Behavioural science shows us that once someone knows where to look in a complex situation, they can’t “un-see” it.  

But explaining expert perception to someone new is remarkably difficult. 

Experience fundamentally changes how people see the world, like an ambiguous image that suddenly resolves once the hidden pattern is revealed. 

 Image Credit: “How Emotions Are Made: The Secret Life of the Brain (2017) by Dr. Lisa Feldman Barrett. 

In complex domains like law, finance, consulting, and public policy, what matters most is not rule-following, but learning by doing in messy, often high-stakes, environments. 

Over time, experts develop pattern recognition and a finely tuned sense of what to attend to. But this knowledge becomes invisible to them. The most valuable insights become instinctive. Senior professionals rarely articulate how they know what they know, because much of that knowledge operates below conscious awareness. 

This creates a structural vulnerability. The expertise organisations value most consists of tactical trade-offs, strategic judgement, and subtle cues built up over years. Yet because this knowledge is rarely documented, firms often don’t realise how much of it they depend on, until it disappears. 

The Quiet Knowledge Crisis Emerging in Professional Services 

Institutional memory erodes not simply because people move on, but because the invisible thinking that made them effective was never captured or transferred in the first place. 

At the same time, firms report growing difficulty finding “experienced” talent. They’re looking for more than just years served. It’s the ability to apply knowledge in context, navigate ambiguity, and make sound decisions under pressure. Raising experience requirements, as some firms are doing, won’t create these capabilities. Instead, it shrinks talent pools without solving the underlying issue. Junior staff need rich opportunities to develop judgement in context.  

In practice, this means firms do not have an experience shortage so much as an experience-creation problem. As traditional career pathways narrow and junior roles are stripped back, organisations are asking for experience without providing the conditions in which it can form. 

The Missing Middle: Where Professional Judgement and Expertise Actually Develop 

Trainees know the theory. Senior professionals can navigate reality. Through years of client work, they have developed the experiential know-how to instinctively weigh strategic trade offs and make decisions.  

What’s disappearing is the bridge between the two: the experiential learning that transforms theoretical knowledge into practical judgement. 

Historically, apprenticeship-style learning closed this gap. Juniors absorbed expertise by sitting near experts, overhearing conversations, watching decisions unfold, and learning how strategies evolved in real time. Crucially, the “learning by osmosis” model transmitted not just knowledge, but ways of thinking. That model is breaking down. 

Hybrid working and automation have dramatically reduced exposure to expert reasoning. Many juniors now see the outputs of decisions without ever witnessing the thought process behind them. 

As AI compresses traditional career ladders, firms can no longer rely on experience emerging naturally over time. Waiting for “ready-made” experience has become both unrealistic and exclusionary. Experience now has to be deliberately created through workflows, roles, and AI systems that expose professionals to judgement, trade-offs, and decision-making in context, rather than shielding them from it. 

Without new ways to surface and transfer this invisible expertise, the capability gap will only widen until we reach the tipping point of irreversible skill decay.  

When AI Replaces Thinking, Professional Capability Decays 

Many professional services firms approach AI as a tool problem: how to train people to use it efficiently so that they can be more productive, provide better client service and ultimately make the firm more money. The appetite for this is clear. A 2025 Thomson Reuters survey found that 55% of professionals report significant changes in how they work due to AI adoption, while 88% said they would favour profession-specific AI assistants.  

Yet, improving tool adoption and proficiency doesn’t solve the growing cognition gap.   

Most AI tools are designed to push information at users rather than developing their thinking capabilities. They provide answers, summaries, and recommendations, but rarely prompt reflection, sense-making, or judgement. While this boosts speed, it risks short-circuiting the cognitive effort through which expertise forms. Professionals may become faster, but not necessarily better. 

This matters because expertise does not develop from exposure to answers alone. It develops through grappling with uncertainty, weighing trade-offs, and understanding why decisions unfold the way they do. 

In 2026, the danger is that technology shortcuts the thinking process so effectively that people stop laying down new knowledge altogether. If AI always decides what matters, professionals never learn to recognise it themselves. 

Outcomes improve when professionals think first and then use technology. Thinking has to come first.  

Why Knowledge Management Systems Fail to Capture Judgment 

Knowledge management systems have become excellent documentation catalogues,  flawlessly organizing the case studies, templates, and playbook that show how to do things.  

Yet, there’s a massive missing data set — the unwritten rules of how work actually gets done. What experts notice. When they change course. Which signals matter and which can be ignored. How trade-offs are navigated when there is no obviously correct answer. This invisible thinking exists in the gap between “work as imagined” and “work as done”. 

Large language models (LLMs) do not contain this knowledge because it’s not documented. It‘s part of the lived experience. And unless organisations find ways to help experts surface it, AI is poised to accelerate its disappearance rather than preserve it. 

From Automation to Cognition Support: Redefining AI’s Role in Professional Services 

In 2026, leading professional services firms will draw a sharp distinction between AI designed to automate tasks and AI that improves cognition. 

Automation-focused AI excels at efficiency. Cognition-focused AI is grounded in behavioural science and designed to surface and enhance judgement, rather than replace it. 

Behavioural science-led AI focuses on better questions instead of faster answers. It prompts professionals to pause and reflect, articulate their reasoning, and think aloud about their work. In doing so, it deepens thinking and surfaces mental models experts didn’t realise they had — and that are so critical to providing the exceptional work that sets firms apart.   

This is especially important for senior professionals, who generally need help identifying the cues and trade-offs they use unconsciously. When their thinking becomes visible to themselves and others, it also becomes transferable. Experts can refine their own reasoning, test assumptions they didn’t know they were making, and continuously sharpen their judgement. This visibility also makes their expertise explainable to clients: strengthening trust, demonstrating value, and improving willingness to pay, and retention. For teammates, it reduces rework and misalignment by clarifying not just what is needed, but why it matters and how decisions should be approached. When expertise is made explicit, it can be organised and shared for the benefit of all teams and clients, current and future.   

Real professional work is not linear. It involves twists, course corrections, and competing priorities. AI systems that respect this complexity, rather than smoothing it away, are the ones that will help organisations preserve and scale expertise, rather than replace it. 

Key Takeaways for Professional Services in 2026 

1. The biggest AI failures will be cognitive, not technical
Firms that focused solely on speed will face skill decay as experiential learning opportunities vanish. This will be a learning failure, not a technology failure.

2. Expertise will become an intentional design opportunity
As automation and hybrid working squeeze out learning opportunities, firms will need to intentionally create micro-opportunities for junior staff to build judgement, reflection, critical thinking and decision-making skills, supported by AI that surfaces and shares expert thinking in context.

3. AI that amplifies human judgement will outperform AI that replaces it

The most valuable AI systems will make invisible expertise visible, creating new “expertise datasets” rooted in how professionals think and reason.

4. The most successful talent strategies will shift from hiring experience to creating it

Firms that focus on helping people build experience will outperform those that simply demand experience upfront. 

The Choice Ahead for Professional Services Firms 

The risk ahead is not whether AI can do the work, but what gets lost when AI makes work look easy and professionals stop learning how to think and make the tough judgement calls.  

Firms that treat AI purely as an efficiency tool will find their expertise quietly eroding, while those that use AI to surface judgement will develop, scale and improve critical thinking,  even as machines and LLMs become more capable. 

When it comes to developing the next generation of professionals to deliver exceptional client outcomes, the differentiator will not be who adopted AI fastest, but who adopted it most intelligently.