Professional services firms increasingly present themselves as technology-driven organizations. Annual reports, strategy presentations, and leadership interviews are filled with references to AI-enabled delivery, integrated knowledge systems, automation platforms, intelligent workflows, and scalable client solutions. The language increasingly resembles the vocabulary of software companies rather than traditional partnerships. Large networks openly discuss moving beyond labor-intensive delivery models while simultaneously investing billions into cloud alliances, offshore delivery centers, centralized platforms, and generative AI capabilities. Microsoft partnerships, AI copilots, workflow orchestration layers, and enterprise-wide knowledge systems increasingly sit at the center of firm strategy rather than somewhere inside the IT department. (Microsoft and PwC Strategic Alliance, EY AI Platform Strategy, Deloitte Generative AI)
Many of these AI ambitions are now being built on top of delivery infrastructures that already became highly centralized over the last decade through global delivery centers, offshore execution hubs, and industrial-scale operational platforms. I explored these operational shifts further in The Silent Engine: How Global Delivery Centers Are Rewiring Professional Services Firms.
Yet underneath the momentum sits a contradiction the industry still struggles to address. Most firms are trying to build platform economics on top of organizational structures originally designed for localized relationship businesses. The technology is evolving rapidly, but the underlying economics of ownership, incentives, revenue attribution, and governance remain deeply tied to the partnership model. And the moment AI systems begin turning expertise into reusable assets instead of engagement-specific delivery, those tensions start surfacing everywhere inside the institution.
Many of these tensions already appear in the industry’s core performance management systems. Traditional partnership KPIs such as utilization, top-line growth, and local bottom-line profitability often reward behavior that conflicts with long-term platform economics, centralized investment, and reusable capability development. These structural incentive conflicts are explored further in The Partnership KPI Trap and The Utilization Trap.
This is why many current discussions around AI in professional services feel strangely incomplete. The challenge is not whether firms can deploy copilots, automate workflows, or build centralized technology platforms. The large firms already know how to do that. They have been operating global technology environments for years. The challenge is that reusable AI fundamentally changes how value is created and distributed across the network. And that becomes politically difficult the moment work starts compounding instead of being repeatedly recreated. This dynamic closely mirrors the tensions explored throughout the broader Professional Services Transformation Paradox series.
The Industry Already Knows How to Centralize Technology
This distinction matters because some observers still underestimate how technologically sophisticated large professional services networks already are. The Big Four have operated centralized audit technology environments for years. Deloitte operates Omnia, EY runs Canvas, KPMG built Clara, and PwC standardized around Aura. These platforms already centralize methodologies, workflows, analytics, testing procedures, and documentation standards across highly complex international networks. (Deloitte Omnia, EY Canvas, KPMG Clara, PwC Aura)
But the direction is no longer limited to the Big Four. Across the broader upper tier of professional services, firms increasingly build centralized audit and delivery environments that resemble technology operating systems more than traditional partnership infrastructure. Grant Thornton developed Leap and increasingly positions GTAP as an AI-enabled audit platform designed to standardize workflows, analytics, and execution across the network. Forvis Mazars built Atlas as part of its globally integrated audit model, reinforcing the firm’s long-standing push toward stronger international consistency and centralized methodologies. Even firms historically organized as looser international networks increasingly require common technology layers to maintain quality control, workflow standardization, analytics capabilities, regulatory consistency, and AI readiness across jurisdictions. (Grant Thornton Leap and GTAP, Forvis Mazars Atlas)
That shift matters because technology standardization quietly changes institutional power structures inside professional services firms. The more delivery depends on centralized platforms, shared AI layers, common data architectures, reusable workflows, and globally governed methodologies, the harder it becomes to operate as a purely decentralized collection of locally optimized partnerships. In many ways, the technology architecture itself is beginning to pull the industry toward more centralized operating models, regardless of whether the formal governance structure officially changes or not.
But maintaining a centralized audit technology stack is fundamentally different from building reusable AI-driven delivery economics across an entire professional services organization.
Audit platforms primarily standardize execution and quality control inside highly regulated delivery models. They do not fundamentally change how profits are distributed across the network. They do not significantly reduce dependency on labor-based economics. And they operate inside comparatively controlled governance structures because audit methodology standardization directly supports regulatory compliance and audit quality objectives.
AI-enabled platform economics are different because they start changing where value actually resides inside the institution. Once intelligence becomes reusable across engagements, service lines, and member firms, the economic center of gravity starts moving away from localized delivery and toward ownership of the underlying systems themselves. That transition creates tensions which traditional centralized audit technology initiatives largely avoided.
The Industry Has Already Tried Parts of This Before
There is also a historical irony embedded in the current AI discussion. Professional services firms have repeatedly attempted to move beyond pure labor-based delivery models over the last two decades. Large firms experimented with software products, managed services, offshore industrialization, subscription models, platform businesses, and reusable technology-enabled solutions long before generative AI became mainstream.
Some initiatives worked operationally while struggling economically inside the partnership structure itself.
Selling software inside partnerships often created uncomfortable questions around investment horizons, ownership rights, cross-border revenue allocation, support models, and incentive systems. Managed services businesses frequently collided with cultures optimized around high-margin advisory work rather than long-term operational delivery. In many cases, firms discovered that building scalable products required fundamentally different governance models, capital structures, pricing logic, and operating disciplines than traditional partnerships were designed for.
AI now pushes those tensions significantly further because the reusable asset is no longer simply software. Increasingly, the reusable asset becomes embedded expertise itself.
And that changes the conversation entirely.
The Real Friction Starts Once Intelligence Becomes Reusable
Traditional professional services economics remained relatively coherent as long as expertise stayed closely connected to localized human delivery. Partners owned relationships. Member firms optimized local profitability. Service lines controlled resources and staffing. Revenue attribution remained comparatively visible because work was still tied directly to identifiable engagements and delivery teams.
Reusable AI systems weaken those boundaries.
Once firms start embedding expertise into shared delivery platforms, workflow engines, continuously improving models, centralized knowledge architectures, or intelligent automation layers, value creation begins decoupling from local execution structures. The first engagement may finance the capability, but subsequent engagements increasingly reuse the underlying intelligence at dramatically lower marginal cost.
At that point, the old economic assumptions start becoming unstable.
Who owns the reusable asset when it gets deployed globally? Which member firm receives economic credit when a centrally trained capability improves delivery margins elsewhere in the network? Which service line funds ongoing development if the primary benefits materialize outside its own P&L? And how should compensation systems behave when value increasingly comes from shared platforms rather than localized execution effort?
Most firms do not yet have convincing answers to these questions.
Instead, many networks are quietly rebuilding similar capabilities repeatedly across different geographies and service lines because the incentive structures still reward local optimization over network-level compounding. From the outside this often appears wasteful. Internally, however, the behavior remains entirely rational within decentralized partnership economics. Similar tensions already emerged in earlier waves of globalization and shared services expansion across professional services firms, where local P&Ls frequently resisted centralized optimization initiatives.
This becomes particularly difficult in firms where partner economics remain heavily tied to local P&Ls, utilization targets, and short-term contribution margin metrics. Under those conditions, investing time into reusable centralized capabilities may reduce local economics in the short term even while improving long-term enterprise value for the wider network. These tensions closely mirror the structural dynamics explored in The Contribution Margin Trap and The Partnership KPI Trap.
Independence Rules Quietly Make the AI Problem Even Harder
The AI discussion becomes even more complicated once audit independence enters the picture.
Unlike many technology companies, professional services firms operate inside highly restrictive regulatory environments around client data usage, independence, confidentiality, and conflicts of interest. Particularly within audit, firms cannot simply aggregate client data freely into centralized commercial AI training environments the way large technology platforms can.
That creates a major structural limitation many external observers still underestimate.
Some of the most valuable datasets inside professional services networks sit inside highly sensitive audit environments. Yet audit independence frameworks severely constrain how that information can be used, shared, monetized, or repurposed. Even where firms possess enormous amounts of operational and financial data, regulatory boundaries often prevent unrestricted reuse for broader commercial AI training purposes. (SEC Auditor Independence Rules, IFAC International Independence Standards)
This creates an asymmetry between professional services firms and native AI platform companies. Technology firms often improve their systems by continuously aggregating and learning from large-scale user interaction data across centralized environments. Professional services firms, particularly in audit, face significantly tighter constraints around how knowledge and data can legally flow across clients, teams, geographies, and commercial contexts.
And the tension becomes even sharper when firms simultaneously operate audit, advisory, tax, and technology businesses under one global brand.
Because now the AI discussion is no longer simply about technology strategy. It becomes intertwined with audit independence, regulatory exposure, data governance, risk management, and institutional trust itself.
AI Quietly Challenges the Leverage Model Underneath the Industry
The deeper issue, however, extends beyond governance and regulation. Reusable AI systems increasingly challenge the leverage structures professional services firms have depended on for decades. Traditional partnerships were built around scalable apprenticeship models where large volumes of junior execution simultaneously generated revenue, trained future talent, and expanded delivery capacity.
AI starts compressing those mechanics.
If significant portions of operational work become automated, fewer junior resources may be required for delivery. Expertise may increasingly accumulate inside centralized systems instead of inside delivery pyramids. Knowledge transfer may happen through workflows and embedded tooling rather than repetitive manual execution. Over time, the relationship between labor input and client value may weaken substantially.
This creates an institutional contradiction many firms still underestimate. The industry wants the productivity benefits associated with reusable intelligence while preserving economic structures originally designed around labor-intensive leverage models. Those two ambitions do not naturally align forever.
In many firms, those economics are still reinforced through utilization-driven management systems that reward billable hours and labor intensity rather than automation, simplification, or reusable delivery. AI therefore creates a direct collision between the industry’s historical leverage economics and the emerging economics of intelligent platforms. In some cases, the existing KPI systems may actively discourage the very productivity gains firms publicly claim to pursue, as explored in The Utilization Trap.
That is one reason many AI initiatives still remain carefully contained despite enormous public enthusiasm. The technological barriers are often manageable. The organizational implications are far harder. The same structural tensions are already visible in broader debates around contribution margins, utilization-driven economics, service delivery centers, and hidden operating costs across professional services networks. Many firms still optimize local contribution margins while large portions of technology, platform, compliance, middle-office, and centralized delivery costs remain structurally disconnected from frontline economics, as explored in The Cost Reality.
Private Equity May Accelerate What Partnerships Struggle to Approve
This is also one reason private equity has become increasingly interested in professional services over the last several years. Most commentary still frames these investments primarily around consolidation, operational efficiency, or access to recurring revenue streams. Those factors matter, but they may not be the most important part of the story anymore. As explored in The Seven Ways Private Equity Is Breaking Into the Big 10, private capital is increasingly entering the sector not simply to finance growth, but to reshape governance structures, delivery models, and the economics of how professional services firms operate.
Private equity changes the economic logic of decision-making inside the institution itself.
Traditional partnerships often struggle to make large centralized investments whose benefits materialize unevenly across the network or only emerge over longer time horizons. Member firms protect local economics. Partners naturally focus on annual profit distributions. Large-scale platform investments frequently create political resistance because they require accepting temporary imbalance between who funds the investment and who ultimately benefits from it. That becomes particularly difficult once AI-enabled systems start shifting value away from localized delivery and toward centralized reusable assets.
This challenge becomes even harder in firms where annual partner compensation, utilization targets, and local bottom-line metrics dominate decision-making. Under these conditions, enterprise-level investments into centralized AI platforms may appear economically irrational at local level even while being strategically necessary for the institution as a whole.
Private equity-backed structures can alter those constraints significantly faster than traditional partnerships typically can. External capital allows firms to fund large technology investments without immediately reducing annual partner distributions. Centralized ownership structures make it easier to enforce network-wide platform decisions. Governance models become more hierarchical. Investment horizons can extend beyond annual compensation cycles. And perhaps most importantly, economic value can increasingly be measured at enterprise level rather than primarily at local partner or member-firm level.
This does not automatically solve the underlying tensions. In some cases, it may intensify them. The pressure for scalability, margin expansion, and centralized control may accelerate far faster than the cultural adaptation inside the firm itself. But private equity-backed firms may nevertheless possess one major structural advantage in the AI era: they can rewrite economic rules faster than traditional partnerships can negotiate them.
That matters because AI-enabled platform economics increasingly reward centralized ownership, reusable intellectual property, standardized delivery layers, and enterprise-wide optimization. These are precisely the areas where decentralized partnerships often encounter political friction. Private equity-backed structures may therefore gain advantages not simply because they have more capital, but because they can move institutional power and economic incentives into alignment more aggressively. Recent examples across the sector include Grant Thornton US, Baker Tilly, Citrin Cooperman, and Interpath Advisory, all of which reflect increasing private equity influence over the future operating models of professional services firms. (Cf. Breaking Partnerships: The Seven Ways Private Equity Is Breaking Into the Big 10)
The implications could become significant over the next decade. The industry may gradually split into two very different models operating under similar brands. One side continues functioning primarily as a decentralized partnership optimized around localized relationships and labor-based economics. The other increasingly behaves like a centrally governed platform business built around reusable systems, managed services, AI-enabled delivery, and enterprise-level economics.
And that may ultimately become the real significance of private equity entering professional services. Not merely as a new source of capital, but as a mechanism for accelerating organizational change that traditional partnership structures may struggle to execute on their own.
Closing Thoughts
The firms that navigate this transition successfully will likely centralize far more control around technology assets, AI governance, pricing structures, workflow architectures, and delivery infrastructure than many partnerships are historically comfortable with. Not because centralization is philosophically superior, but because reusable intelligence naturally concentrates economic value around ownership of the underlying systems.
And that is where the conversation becomes politically sensitive inside many networks.
Because the AI transition is no longer simply a technology modernization effort. It increasingly becomes a negotiation about power, economics, and institutional control inside the professional services model itself.
The irony is difficult to ignore. Professional services firms may become exceptionally good at helping clients build AI-enabled platform businesses while remaining structurally constrained from becoming platform businesses themselves. Not because they lack technical capability, but because the economics of decentralized partnerships often resist the very type of centralized compounding systems AI naturally favors.
And that may ultimately become the defining paradox of the industry’s AI era.
What This Means for Boards
Boards should avoid viewing AI inside professional-services firms primarily as a technology initiative. The deeper challenge is structural. AI increasingly changes where value is created, how work is delivered, how leverage models function, and where operational dependency accumulates.
This creates tensions that traditional partnership models were not originally designed to manage. Centralized AI platforms require large-scale investment, integrated data environments, standardized workflows, and governance structures capable of coordinating decisions across multiple firms, service lines, and jurisdictions. At the same time, many professional-services firms still operate through highly decentralized economic and political structures.
Boards should therefore look beyond AI capability demonstrations and focus on the underlying operating model implications:
- Who controls the AI platform?
- Who funds the investment?
- How are benefits distributed across the network?
- What happens to the traditional leverage pyramid?
- How does AI affect partner incentives, talent development, and delivery economics?
- Where does operational dependency accumulate over time?
The firms that manage these tensions deliberately may strengthen their long-term strategic position significantly. The firms that do not risk creating growing disconnects between centralized technology infrastructure and decentralized partnership governance.
Continue Exploring
The Bigger Picture
The Professional Services Transformation Theory
Why artificial intelligence, changing economics, new operating models and growing investment requirements are part of a broader structural transformation of professional-services firms.
Explore the Economics Underneath AI
The AI Cost Stack: Why Professional Services Firms Are Looking for AI Savings in the Wrong Place
Why AI may reduce visible delivery effort while simultaneously increasing the technology, orchestration, governance and capital required underneath professional work.
Explore How Expertise Is Changing
When Expertise Becomes Metered Infrastructure: How AI Is Changing the Economics of Professional Services
How artificial intelligence is beginning to separate expertise from the individual professional and turn parts of professional knowledge into reusable infrastructure with fundamentally different economics.
Explore the Organisational Consequence
The Two-Speed Firm: Why Professional Services Firms Are Quietly Splitting Into Multiple Economic Systems Under One Brand
Why centralized platforms, AI-enabled capabilities and industrialized delivery increasingly operate according to different economics from traditional partner-led businesses, creating competing operating logics inside the same institution.
Apply the Thinking to Your Firm
AI Economics Review
An independent review of what AI is actually changing in your firm’s economics, including client demand, delivery economics, pricing, value capture, investment requirements and the wider operating model.
Sources
Primary Sources
- Deloitte Omnia
- EY Canvas
- KPMG Clara
- PwC Aura
- Grant Thornton Leap
- GTAP
- Forvis Mazars Atlas
- Microsoft and PwC Strategic Alliance
- EY AI Platform Strategy
- Deloitte Generative AI
- SEC Auditor Independence Rules
- IFAC International Independence Standards
- Grant Thornton US PE Transaction
- Reuters on Baker Tilly and Moss Adams Combination
- Bridgepoint Investment in Interpath Advisory
Secondary Sources
- The Partnership KPI Trap: How Top Line, Bottom Line, and Utilization Can Push Firms Against Their Own Strategy
- The Utilization Trap: Why Professional Services Firms Are Optimizing the Wrong Productivity Metric
- The Contribution Margin Trap: Why Professional Services Firms Are Optimizing the Wrong Economics
- The Cost Reality: Why Front, Middle, and Back Office Economics Don’t Add Up
- The Exit Problem: Private Equity Has Found Ways Into Professional Services. Getting Out Is Harder
- Case Study 27: Baker Tilly and Private Equity
- Case Study 28: Forvis Mazars — One Brand, Two Firms
- The Professional Services Transformation Paradox Series