PwC

PwC (PricewaterhouseCoopers) is one of the Big Four professional-services networks, with audit, tax, and advisory practices in more than 140 countries and roughly 370,000 people globally. PwC International Limited coordinates a network of legally separate member firms. The US firm is the largest member firm by revenue and the locus of the May 5 OpenAI and May 14 Anthropic announcements relevant to this site.

Where this came from

PwC entered this cycle through two linked disclosures in the same month. The May 5 OpenAI release positioned PwC as customer-zero for an internal finance-procurement agent. The May 14 Anthropic expansion positioned PwC as alliance distributor with named client references.

What the expectation was

For a both-lane company, the expectation is explicit evidence separation. Buyer lane should show internal deployment outcomes with before-and-after metrics. Provider lane should show carry-over limits, pricing, and named-client outcomes that are not confounded with alliance marketing language.

Where it succeeded

PwC succeeded in making the dual role visible. It provided enough signal to map a buyer lane and a provider lane without forcing a false single narrative.

Where it failed

The disclosures still under-deliver on quantified outcomes in both lanes. Internal buyer metrics are limited, and provider carry-over remains partly inferred from alliance framing rather than independently disclosed operating evidence.

Where it stands now

Current posture is active with guarded confidence. Keep PwC as a strategic both-lane anchor, but keep funding-defense claims on hold until buyer and provider evidence are separated with stronger measured outcomes.

Lever is clear: 1 of 2

The mixed-lever posture is credible, but lane-specific lever attribution is still partial.

Human-side evidence: 1 of 2

Named leaders are on record, but operator-level workflow outcomes remain thin.

Proof strength: 1 of 2

Evidence is still primarily T2 and announcement-driven.

Decision usefulness: 2 of 2

High strategic utility as a both-lane governance case, provided evidence is split by lane in decision discussions.

Total: 5 of 8

Current decision: hold for broad funding acceleration, proceed for controlled pilot design and lane-specific audits.

What would change this: disclose one internal buyer before-and-after metric and one externally validated client outcome tied to a named alliance workflow.

The 2026 AI posture deliberately runs on two model channels: OpenAI for the internal CFO-function customer-zero work (Native Finance Function, May 5), and Anthropic for broad practice-wide deployment across technology, deal execution, and client engagement (May 14 alliance expansion). The customer-experience implication is dual: PwC's own finance organisation pilots agents that may later ship to clients, and PwC's advisory engagements increasingly include Claude-backed delivery. The May 14 announcement also named Advocate Health as a client, providing an early buyer reference inside the same release.

Audit, tax (including managed tax services), and advisory (deal advisory, consulting, risk, technology transformation). Sector practices span financial services, health industries, energy, technology, consumer markets, and government. The Native Finance Function (with OpenAI) and the Claude-based delivery patterns under the Anthropic alliance are the headline 2026 AI artefacts.

Paul Griggs is US Senior Partner and is the named voice on the May 14 Anthropic alliance expansion. Tyson Cornell is US Advisory Leader and is the named voice on the May 5 OpenAI Native Finance Function announcement. Mohamed Kande is Global Chair.

The other three of the Big Four: Deloitte, KPMG, EY, each running its own frontier-lab strategy. Strategy consultancies (McKinsey, Bain, BCG) compete on the advisory layer. Technology-implementation specialists (Accenture, Capgemini, Infosys, TCS, Wipro) compete on systems integration. Audit competition is concentrated among the Big Four themselves under regulatory rotation rules.