For centuries, apprenticeships have been used to train workers in a myriad of fields. The premise is fairly simple: Younger workers are paired with experienced masters in their respective industries to learn how to do a trade the right way.
Switzerland has long championed this model; more than two-thirds of Swiss students choose an apprenticeship after completing compulsory schooling, rather than proceeding directly to higher education. Companies benefit from a steady pipeline of new talent, reduced recruitment costs, and a boost in morale as they support employee growth. This dual-system approach also contributes to Switzerland’s low youth unemployment rate and allows lifelong upskilling of the existing workforce.
While apprenticeships are often associated with manual trades globally, the Swiss model has long successfully covered complex service sectors, from banking to administration. Now, Geneva-based DiploFoundation – a nonprofit focused on diplomacy, governance and digital policy – is extending this proven methodology to the cognitive realm of AI.
By incorporating the ‘learn-by-doing’ approach to AI, workers are gaining a deeper understanding of the technology behind today’s AI models and the most effective ways to deploy these tools in their professions. Building on that long-standing tradition, this model offers a compelling blueprint for nations and institutions seeking to develop an AI-literate workforce.
What is an AI Apprenticeship?
Diplo’s model of AI apprenticeship is an immersive, several-week training course where participants work directly within AI systems to learn the logic behind today’s large language models (LLMs) and how to build AI agents themselves. Our ultimate goal is to ensure that all participants can accurately explain the technology in a way that a child can understand, because we believe that if you can’t do so, then you aren’t truly an expert.
We are able to accomplish a comprehensive understanding through practical, project-based assignments paired with direct guidance, mentorship and structured educational support. Rather than attending lectures about AI or listening to sales pitches about AI models, our apprentices are developing and deploying their own AI tools to solve problems in their work.
The Swiss model has long successfully covered complex service sectors, from banking to administration.
While anyone can build a bot in five minutes, deeply understanding its logic requires weeks of friction and practice. This empowers individuals not only with a fundamental understanding of how AI works but also with the skills needed to integrate it into their own lives and jobs, whether they continue on to study and work in the field or not.
AI apprentices often lack a technology background. In fact, we have found that individuals without tech backgrounds make for stronger participants. Traditional programmers are trained for deterministic systems, where code executes logic in a linear manner.
AI, however, is a probabilistic system. It requires a different cognitive approach, akin to a philosophical inquiry or Socratic dialogue, where the user negotiates meaning with the machine. We find that humanists often grasp this nuance faster than engineers.
Why others should consider adopting this model
The reasons for the broad, global adoption of the AI apprenticeship model include the following:
- Bridging the AI-skills gap: Many countries face a shortage of AI-capable talent. Similar to the early iterations of the internet, capacity development (or maybe even ‘vocational training’) is needed to bring active workers up to speed on the capabilities and limitations of the new technology. The apprenticeship model offers a scalable route: By embedding learning in workplaces and real problems, it accelerates skill development and ties into employer demand.
- Embedding legitimacy and governance from the start: As AI becomes integral to public services, diplomacy, business and civil society, the risks of deploying opaque or ungoverned models grow. Through apprenticeship frameworks, participants gain not only technical skills for building models but also an understanding of the principles of legitimacy, ethics and governance. This prepares them to decide when and how a model should be deployed and to critically evaluate potential AI solutions.
- Adaptability: For a field changing as rapidly as AI, the apprenticeship model offers a distinct advantage: It is built to be adaptive. Curricula can be reviewed regularly in collaboration with employers and trade associations that support the program. Working directly with experts in real-world settings ensures that the material remains relevant to current tools, challenges and use cases. In a domain where both technology and governance evolve rapidly, a flexible and continuously updated education model is essential for long-term sustainability and value.
- Economic and social resilience: As AI evolves, it’s critical that the workforce of tomorrow not only knows how to prompt a system, but how to shape AI to solve society’s most complex problems.
A workforce that is truly competent in AI will be better positioned in an economy transformed by automation, AI integration and data-driven decision-making. Looking ahead, it’s clear that regions without an AI-literate workforce will quickly fall behind those that prioritize talent development.
By investing in apprenticeship-based AI training, communities can build local capacity, strengthen innovation ecosystems, reduce brain drain and enable high-value activities to grow where they originate.
Drawing on Switzerland’s centuries-old vocational training model, AI apprenticeships present a blueprint for developing an AI-literate workforce – one that not only knows how to use AI tools but also understands how the systems function and can critically engage with, govern and scale the technology. For nations, educational institutions and employers navigating the AI era, apprenticeship-style training is a smart, future-focused investment.
