Anthropic, the company behind Claude, wants to train the engineers who will put its AI to work inside organizations. On October 2, 2026, it announced Claude Frontier Academy, a training program backed by a US$100 million commitment, with a target of training 10,000 engineers by the end of 2027.
For current plans, prices and every Maitz article on Claude, see our Claude hub.
Their job will be to turn a potential AI use case into a system a business can actually run: choose a worthwhile problem, connect Claude to the organization’s systems and information, complete security reviews, and hand the solution over to the team that will use it. Anthropic calls the role a Frontier Deployed Engineer. Engineers from Accenture, Bain, Deloitte and Morgan Stanley are among the participants in the first cohorts.
Four days in person, followed by a workplace project
According to the program details, training begins with four days in person. During the first three days, participants build a Claude system for a fictional organization, working from requirements through to handover. On the fourth day, they are assessed on a new scenario.
Those who pass move on to lead a Claude project in their own organization for 12 weeks, with support from Anthropic engineers and learning alongside the other participants. A further practical assessment is required for final certification. The company expects to award the first certifications in early 2027.
How do engineers join Claude Frontier Academy?
Initially, the program is intended for Anthropic’s enterprise customers and partners, with candidates nominated through their organization. Cohorts are already running in San Francisco, New York and London. Organizations interested in joining can check eligibility with their Anthropic account team or partner manager.
Candidates are expected to have a strong software engineering background and experience building with language models, the AI systems behind services such as Claude. For people who want to begin learning independently, the program page also points to a separate offering, Claude Academy, which is open to everyone.
The workplace project is the most interesting part of the approach: it puts the emphasis on implementing AI in a real process, beyond becoming familiar with the tool. That is also where the program will be tested. The figure of 10,000 engineers is a target for now, and the value of the training will become clearer when the systems they build meet everyday work.
Sources
- Anthropic: Claude Frontier Academy announcement
- Claude: Frontier Academy program details
