Long before Anthropic announced its new academy, employers were already describing a more mundane obstacle to AI adoption than model capability: a shortage of people able to put the technology to work. The OECD has found that skills gaps limit uptake, particularly among smaller businesses; 40% of manufacturing and finance employers cite skills as their main barrier, while more than half of SMEs are not using generative AI. It argues training should be paired with transparency, accountability, safety and worker-privacy protections.¹

Anthropic is now placing a $100 million bet on that diagnosis. The company’s Claude Frontier Academy aims to train 10,000 Frontier Deployed Engineers by the end of 2027, starting with engineers from Accenture, Bain, Capgemini, Commonwealth Bank of Australia, Deloitte, McKinsey, Morgan Stanley and Novo Nordisk. Its premise is blunt: enterprises do not merely need access to Claude; they need people who understand their own operations well enough to carry an AI project through security review, deployment and adoption.²

Steve Corfield, Anthropic’s global head of business development and partnerships, says “a small team of high-agency people” with the right skills and business knowledge “can transform an entire company.” The harder task, he says, is developing engineers capable of taking Claude “from an idea to a system in production.”²

The first residency borrows from medical training: candidates nominated by their organisations complete an in-person simulated deployment, then lead a real 12-week Claude use case before a final assessment. For Accenture, that reflects a familiar discipline: “judgement tested against a simulated enterprise deployment, not technical skills in isolation.”¹

Novo frames the stakes more concretely. Its data and AI chief, Loic Giraud, says Claude is already touching research and drug discovery, and argues that pairing frontier models with scientific expertise could help bring medicines to patients faster.¹

The shared view is that AI’s next contest will not be won in a demo. Anthropic’s partners want trained operators who can make systems work inside governed, complicated organisations; the OECD’s warning is that those gains will remain uneven unless the broader workforce can follow.