A conversation shaped by voices on stage and votes in the room
When code competes with talent and capital—what happens?
That question sat at the center of our recent webinar, which brought together nearly 100 professionals across 4 regions and 26 countries. While the media has been quick to call AI a bubble—MIT’s much-quoted line that “95% of pilots fail” dominated headlines this summer—our discussion painted a very different picture.
When asked where they see AI heading, 82% of our audience voted for exponential growth. That single number set the tone: firms don’t view AI as passing hype; they see it as the curve they cannot afford to miss.
Busting the Myth: AI Isn’t Dead
The MIT study often cited in headlines tells only half the story. Yes, 95% of GenAI pilots struggled to show measurable ROI, but technology wasn’t the culprit. Failures came from three organizational gaps: lack of contextual learning, attempts to bolt AI onto broken workflows, and resistance at the leadership level.
As Stephen Heathcote observed during our panel:
“AI is already embedded in many services… it’s now part of the plumbing in many of our firms. We’ve moved to productivity very, very quickly.”
The biggest misconception is reducing AI to “just another piece of software.” In reality, it represents adaptive intelligence. What some dismiss as “hallucinations” are actually signs of creative problem-solving—the very capability that allows AI to compose, design, and solve problems beyond the reach of traditional algorithms.
The New Operational Model for CPA Firms
Our polling revealed a subtle but important shift in mindset. While earlier surveys showed 99% favoring “people” as the main growth driver, this session saw 16% choosing “code.” Small, but telling. Naval Ravikant’s leverage hierarchy explains why. Labor demands constant management, capital requires access and permission, but code—once written—scales indefinitely. For accounting firms, this poses a strategic choice: stick with traditional people-driven models or embrace AI-native operations that deliver comparable services at vastly different cost structures.
Blueprint for AI-Native Operations
We call the shift from disconnected tools to unified intelligence the “AI Brain Frame.” Building it means integrating four pillars:
> Data warehouse connectivity
> Multimodal AI frameworks with feedback loops to limit hallucinations
> Robust memory systems for continuous learning
> Micro knowledge workers designed for specialized tasks
The breakthrough isn’t in trying to build one AI for an entire practice, but in creating micro-specialists for sampling, reconciliation, checklist automation, and risk scoring—pieces that can then be combined into firm-wide systems.
Four Tactical Pillars of Transformation
I. Talent & Roles
Mary Richter stressed the need to begin with people:
“Talent and roles must be a starting point. Firms need to redefine their roles to explicitly include working with AI.”
Routine work—analysis, compliance, drafting, research—can be AI-powered, freeing professionals for what they do best: judgment, oversight, and client engagement.
II. Processes & Data Governance
Steve Perkins addressed the real bottleneck:
“The elephant in the room is data—good, clean data. You only get there through governance.”
Without clear ownership, stewardship, definitions, and controls, output (including AI results) remains unreliable.
III. Business Models & Pricing
AI-driven efficiency makes hourly billing unsustainable. Firms must pivot to value-based pricing, outcome delivery, and subscription approaches that align incentives with efficiency gains.
IV. Partnerships & Investments
As Heathcote emphasized:
“Partnerships—taking a partnering approach to implementation of AI—is absolutely essential.”
Equally, firms must hold vendors accountable to prove ROI and ensure solutions truly deliver.
AI-native firms are emerging fast—operating leaner, scaling quicker, and embedding AI from the ground up. Unlike past tech shifts, there’s little adjustment grace period; native adoption compresses the timeline, leaving established firms at risk if they delay.
Mary Richter captured it clearly:
“Change must come from the top. If leaders don’t demonstrate adoption, no one else will.”
MIT’s findings underline this: leadership resistance—not the tech—was a key failure driver. Success requires leaders to personally use AI tools, back adoption with resources, and align culture and performance with AI goals.
Our profession faces a stark choice: embrace AI as a transformation driver or risk becoming irrelevant as AI-native competitors surge ahead. The window for proactive change is shrinking. The firms that will thrive are those that stop experimenting and start systematizing—building operations that free their people for human strengths: relationships, judgment, and strategy. When 82% of global practitioners predict AI’s exponential growth, the signal is undeniable. The question is no longer if AI will transform professional services, but who will lead that transformation. Code is already writing its way into the future. The real decision: will you help write it, or be written over?
