TL;DR

Anthropic has launched Claude Frontier Academy, backed by what it calls a $100 million commitment, to train 10,000 “Frontier Deployed Engineers” by the close of 2027. The company announced it on Saturday 3 October. The first cohorts, nominated by firms including Accenture, Deloitte, McKinsey, Capgemini and Novo Nordisk, are already running, one of them in London.

How the programme works

The Academy’s first course is a residency that Anthropic compares to medical training: learn from practitioners, work on realistic cases, then be assessed before working alone. It is aimed at experienced software engineers who have built with large language models and helped colleagues adopt AI. Each organisation nominates its engineers, and every nominee brings a specific Claude project they will run once back at work.

It runs in two stages. A multi-day in-person course, taught by Anthropic staff alongside licensed trainers, takes engineers through a mock company rollout, from picking the use case to security review and handover, and ends with a graded practical. Passing earns a Claude Resident Engineer badge. Engineers then spend 12 weeks leading a real deployment at their own employer with Anthropic support, and a second assessment decides who earns the full Frontier Deployed Engineer badge. Anthropic expects the first of those in early 2027.

Places are by nomination only. To check eligibility, firms are told to speak to their Anthropic account contacts.

Why the consultancies are lining up

The named first-cohort members lean heavily towards professional services, alongside banks Morgan Stanley and Commonwealth Bank of Australia. Anthropic’s pitch is that engineers who can carry AI into production inside a real business are now the hardest people to hire. Steve Corfield, who leads partnerships and business development for Anthropic worldwide, said the hard part is “building that depth of talent”.

The Academy sits on top of Anthropic’s existing partner network, where, by the company’s count, people at 46,000 firms have earned over 175,000 Claude certifications.

For UK readers, the London cohort means the programme is running here from the start, not only in the US. It also follows Barclays’ plan to put half its developers on Claude Code by the end of 2026.

Looking forward

All figures here are Anthropic’s own and the programme has yet to produce a graduate. In our view, the more telling question is what it does to the consulting market: if the big firms’ engineers carry an Anthropic credential, clients choosing an integrator may also be choosing a model vendor. UK buyers commissioning AI work from these firms should ask whether the advice they receive is model-neutral.