Problem
Three years into the generative AI era, most professional services firms have moved past the question of whether to use AI. The tools are bought. The logins are issued. Junior staff are drafting with ChatGPT, partners are experimenting with Copilot, and every quarterly review includes a slide on “AI productivity.” And yet, in boardrooms from mid-sized law firms to regional accounting practices, the same uncomfortable question keeps surfacing: why isn’t this showing up in our margins yet?
The honest answer is that adoption and mastery are two completely different things, and in 2026, the gap between them has become the single largest source of competitive risk in the professional services sector. Most firms have done the easy work (licenses, prompts, pilots). Almost none have done the hard work (skills, governance, workflows, culture). The firms that are pulling away from their peers right now aren’t the ones with the most AI seats. They’re the ones whose people have actually been trained to use them well.
Why It Matters
The scale of the gap is hard to overstate. Skillsoft’s 2026 Workforce Readiness Report found that nearly 9 in 10 workers are now using AI in some form, but most companies still lack the governance, skills visibility, and timely training required to make that usage safe or productive. McKinsey’s 2025 global survey reinforced the point: while AI use is now nearly universal, only a tiny fraction of organizations describe themselves as “mature” in their AI deployment, and the biggest blocker they cite is talent and process, not technology.
The cost of staying on the wrong side of that gap is compounding. McKinsey’s workplace research in 2025 found that 46% of leaders name talent skill gaps as a major barrier to AI adoption. Deloitte’s 2026 State of AI in the Enterprise report tracked a widening divide between organizations reporting “meaningful” value from AI and those still running disconnected pilots. The roughly 1% of companies that have reached what McKinsey calls “AI maturity” are capturing disproportionate financial returns, while the long tail of “adopters” are quietly watching their tool spend grow faster than their revenue.
For small and mid-sized professional services firms specifically, this is a once-in-a-generation sorting event. The Big Four and Am Law 50 firms have the capital to absorb a multi-year transition. A 40-person accounting practice or a 25-attorney boutique does not. The decisions you make in the next 12 months about how your people learn AI will determine whether your firm is one of the winners, or one of the firms being quietly outcompeted by a more disciplined rival down the street.
The AI Approach
Closing the adoption-to-mastery gap is not a tooling problem. It’s a workforce problem. The firms making the leap are treating AI readiness with the same seriousness they once applied to a new case management system, a new accounting standard, or a new partner class. That seriousness shows up in three concrete commitments.
First, assess where you actually are on a real maturity model, and stop pretending experimentation is progress. Microsoft has published a five-stage AI maturity framework (from initial exploration through AI-driven enterprise) that maps capabilities to observable behaviors, not vibes. The MIT CISR Enterprise AI Maturity Model offers a similar four-stage lens focused on operating model, governance, and data. Whichever model you choose, the discipline is the same: have an honest conversation about which stage your firm is in, and what the next stage requires. Most firms we work with discover they’re stuck at stage two, running disconnected pilots, while believing they’re at stage three. That gap is where the value leaks.
Second, replace ad-hoc experimentation with structured, role-specific training. The most successful 2025-2026 professional services training programs are not generic “AI for business” webinars. They are job-embedded curricula that build muscle memory in the actual workflows your people perform. PwC’s US firm, for example, launched the Learning Collective specifically for the AI age, combining essential human skills with AI capability, and began piloting AI immersion sessions with new tax associates in late 2025 ahead of a full rollout. EY’s AI Now 2.0 program analyzes each employee’s role and identifies the specific future skills they need to develop, reportedly already engaging half of its global workforce. The pattern is unmistakable: the firms investing in role-specific AI fluency, not generic AI awareness, are the ones pulling ahead.
Third, build governance and workflow redesign into the same program as training. This is where most firms fail. They buy tools, they train people, and they never change the workflow, so trained people revert to old habits the moment they’re busy. The MIT CISR research is unambiguous: the organizations that reach AI maturity are the ones that redesign their operating model alongside the technology rollout. For a law firm, that might mean rewriting the document review process so AI-drafted memos have a defined human-in-the-loop checkpoint. For an accounting practice, it might mean restructuring the audit workflow so AI handles the transaction testing while humans own the judgment calls. Training without workflow change is, at best, a delayed waste.
The compound effect of these three commitments is what separates the masters from the adopters. A master firm in 2026 doesn’t just use AI. It has a documented maturity level, a trained workforce matched to specific roles, and a redesigned operating model that makes the technology compound. The adopter firm has logins, a Slack channel, and a slow leak of value.
Real-World Examples
Microsoft: Microsoft’s own IT organization has published a detailed walkthrough of how it moved itself through five stages of AI maturity, from “AI aware” to “AI driven.” The progression is less about buying new tools and more about the operating-model changes that let tools actually change how work gets done. For a professional services firm, the takeaway isn’t to copy Microsoft. It’s to recognize that even one of the most resourced AI buyers in the world treats maturity as a multi-year operating-model project, not a software purchase.
PwC US: The firm launched the Learning Collective as a dedicated ecosystem for accelerated growth in the AI age, and separately began piloting AI immersion sessions for new tax associates ahead of a full US rollout. The strategic intent is explicit: prepare juniors for managerial roles from day one, because AI is automating the rote tasks that used to define early-career work. This is a textbook example of a firm treating AI training as workforce redesign, not employee perk.
EY: EY’s AI Now 2.0 program uses AI itself to analyze employee roles and surface the future skills each person needs. Engagement is reportedly already at half the global workforce, making it one of the largest active AI reskilling programs at a professional services firm in 2026. The lesson for smaller firms: even without a custom platform, the principle is replicable with off-the-shelf tools.
McKinsey’s “Superagency” finding: McKinsey’s 2025 research concluded that the productivity gains from AI stick only when accompanied by a human relationship and judgment layer. In professional services, where the deliverable is advice, this insight is the whole game. The firms treating AI as a way to do the same work faster are harvesting one-time gains. The firms freeing humans for higher-judgment client work are building compounding returns.
Action Steps
- Pick a maturity model (Microsoft’s five-stage, MIT CISR’s four-stage, or IDC’s AI-Fueled Organization benchmark) and honestly score your firm this quarter. Write the score down. Share it with your leadership team.
- Audit the last five client engagements for AI touchpoints. Identify the two workflows where AI is currently adding the most time back, and the two where it’s adding risk you haven’t governed.
- Replace one “AI tools” line item with a “role-specific AI training” line item in next year’s budget. Allocate at least 8 to 12 hours per FTE annually for job-embedded AI learning, not generic webinars.
- Pick one workflow (e.g., first-draft document review, transaction testing, client intake) and redesign it end-to-end so AI handles the substrate and a named human owns the judgment. Measure the before/after on both cycle time and error rate.
- Name an AI accountability owner at the partner level. Pilots are owned by enthusiasts. Mastery is owned by someone with budget and authority. McKinsey’s 2025 research is blunt on this: without a named production owner, your pilots will stay pilots.
Call to Action
AI can help your firm keep pace with growth, or it can quietly widen the gap between the tools you bought and the value they actually deliver. Schedule a free 30-minute consulting call, and we’ll map where your firm sits on the adoption-to-mastery curve and what the next stage of implementing AI actually requires.