Accenture Is Certifying 50,000 People on Claude. Anthropic Just Gave Them Four Tracks to Do It.

Accenture has committed to certifying 50,000 of its people on Claude. PwC is targeting 30,000, Deloitte 15,000. Anthropic’s July 23 announcement expanding its Claude Certification Program from one credential to four — Associate: Foundations, Developer: Foundations, Architect: Professional, and Architect: Foundations — comes with the numbers to back up why the Big Three are moving at that scale: since the program launched in March, more than 36,000 consultants across 1,300-plus organizations have already earned a credential, and 400,000 people have completed training through the Anthropic Partner Academy.

The four-tier structure matters more than the headline count. It splits consulting-relevant training (Associate: everyday usage for project leads and mixed business/technical roles) from engineering training (Developer: API, tool use, agent development) and from governance training (Architect: Professional, for enterprise-scale integration and evaluation). That’s a direct answer to a gap BCG quantified back in July: in its AI Radar 2026 survey of roughly 2,400 executives, BCG found that while 35% of companies already use agentic AI and another 44% plan to, only 15% of CEOs qualify as “trailblazers” — most of the gap between adoption intent and adoption results isn’t strategy, it’s people who know how to run the thing.

Steve Corfield, Anthropic’s head of global business development, put the stakes plainly: “Companies pulling ahead will have enough skilled people to put Claude to work on their hardest problems.” For a consulting firm evaluating its own AI-delivery capability, that’s the practical takeaway — a client roadmap built on tools nobody on staff is certified to operate isn’t a roadmap, it’s a wish list. Proctored, identity-verified credentials, delivered at Accenture-scale volume, are Anthropic’s bet that the certification itself becomes the procurement signal.