I read the recent headlines about AI automating throughout our organizations with professional curiosity. As CTIO of PwC France, I do not have the luxury, nor the need of reacting emotionally to those statements. My responsibility is to translate them into operating reality. What changes in a professional services firm when intelligence and automation become entangled and embedded in the workflow? What moves, what holds, and what must be redesigned with intent?
Last week, PwC introduced PwC One globally: a significant milestone in our AI‑first transformation and our vision for the future of professional services. The launch crystallizes what many of us have been building toward: an AI‑enabled environment that brings together PwC’s institutional knowledge, methodologies, and emerging autonomous capabilities to help clients navigate complexity with greater speed and precision. It also places the questions I have been examining into sharper relief. When AI becomes central to how work gets done, the partnership model itself must evolve.
PwC is built on a partnership model grounded in leverage. Associates handle structured analytical work. Managers coordinate and review. Senior managers and partners arbitrate complexity and assume accountability. The pyramid reflects economic logic shaped over decades.
When AI absorbs part of the structured analytical layer, the real issue is the leverage equation. If an associate analyses contracts faster, prepares working papers more efficiently, or synthesizes regulatory updates with AI support, throughput per individual increases. That improves responsiveness to clients and reduces internal friction. It also places pressure on the traditional pyramid. The ratio between junior and senior profiles was calibrated for a world where manual synthesis consumed time. When that time compresses, the structure must evolve.
This shift is already visible in audit preparation, tax documentation, and advisory research. PwC One embeds AI directly into methodologies and professional workflows, extending the reach of expertise while keeping human judgment firmly at the centre of every critical decision. The platform brings together proprietary methodologies, compliance frameworks, and domain expertise with advanced AI capabilities designed to surface patterns, test assumptions, and accelerate analysis. Tax analysis, financial reporting, sustainability assurance, deal diligence, and operating‑model transformation are among the initial areas where these capabilities are being deployed.
Two areas require simultaneous attention: margin and mentorship. Professional services firms balance quality, cost, and development. Junior professionals build expertise through repetition and exposure. If AI absorbs part of that repetition, the learning journey must be redesigned deliberately. Short‑term efficiency gains should strengthen long‑term capability, not weaken it.
Mohamed Kande, our Global Chairman, put it plainly at Davos: the traditional apprenticeship model -where entry‑level employees learn by doing basic tasks- is being disrupted by AI. When structured work accelerates, we must ask: where does learning happen? How do we preserve the development path that builds judgment, not just execution speed?
Clients engage us for insight, reliability, and trust. When AI enables more efficient delivery, the value conversation shifts. Productivity gains must translate into deeper analysis, stronger risk assessment, and better client engagement. The partnership model thrives when trust strengthens alongside efficiency. PwC One reflects this transition: moving beyond episodic projects toward more continuous insight, faster learning cycles, and earlier visibility into risk and opportunity.
The pricing model is evolving in parallel. As Paul Griggs, PwC US CEO, noted in an article with the FT the firm is moving “more and more of our work to outcomes pricing”. PwC One operates on a subscription or consumption-based model where clients access automated services -including M&A due diligence, tax‑related insights, and anomaly detection for sustainability data- without requiring a PwC professional in the loop for every interaction. This is a structural change. When billable hours compress and value delivery accelerates, the economics of the pyramid shift. Revenue becomes less about time multiplied by rate and more about access, insight, and outcomes.
Governance is equally central. Every technology decision in our environment passes through regulatory scrutiny and reputational risk assessment. Audit standards, independence rules, data protection laws, and evolving European AI regulation frame our deployment choices. Embedding AI into core workflows requires traceability, explainability, and disciplined change management. If a conclusion cannot be defended, it does not belong in an audit file. PwC One is built on secure infrastructure and aligned with rigorous data privacy, security, and regulatory standards precisely because trust is non‑negotiable. Management itself evolves under these conditions. Historically, part of management involved coordinating information flow and supervising production. As AI reduces coordination friction and automates reporting layers, managerial value shifts toward judgment under uncertainty, conflict resolution between risk and opportunity, and ownership of consequential decisions.
That responsibility intensifies. Information asymmetry within organizations decreases when knowledge becomes searchable and synthesized at scale. Authority can no longer rely primarily on privileged access to data. It rests on clarity of reasoning and accountability. In a partnership environment, credibility becomes even more visible. Leaders must integrate AI-enabled analysis into coherent, defensible positions for clients and regulators.
Paul Griggs has been explicit about the expectations this creates: partners who are not “paranoid about being AI‑first” will have no place at the firm. If you do not understand how AI reshapes delivery, pricing, mentorship, and governance, you cannot lead within this model. The choice is clear: integrate AI into how you think about leverage and value, or step aside.
From my perspective, the main risk is organisational drift. Tools deployed without operating‑model redesign. Throughput increased without recalibrating performance metrics. Production accelerated without reinforcing quality and mentorship. Subscription models launched without rethinking how professionals develop expertise when routine work disappears.
AI is already embedded in professional services. The essential question is how to align it with partnership economics while preserving the foundation of trust. That alignment requires discipline and architectural thinking. I focus on reshaping leverage responsibly. In a firm built on expertise and accountability, structure is strategy, and technology always reshapes structure.
The era of waiting for answers is over. The era of building new operating models has begun.
