Over the last five years, we’ve invested more than $150 million to help people learn digital skills. We’ve worked with 70 organizations across 41 European countries, reaching millions of people. Our goal is to bridge digital skills gaps for underserved communities.
Now, as we all look toward the AI era, we’re sharing what we learned in a new report. We found four key lessons from our digital skills programs that are shaping our new AI-focused efforts, like the AI Opportunity Fund, which supports the newly announced AI Works for Europe.
1. Solve for context
Every learner is different, and so is every community. The best results come from programs and support designed for their specific needs. For example, Google.org-supported research by Generation identified AI-driven hiring bias against mid-career workers; resulting in the development of a targeted program with an 83% job placement rate.
Wraparound support — like help with living costs and technology access — ensures all students can succeed, regardless of their background. We have tested how to offer this support at scale to reduce the cost per learner. When sisters Asmaan and Farzaneh arrived in Germany from Afghanistan with no digital experience, DigiCo provided the technical and language skills needed to progress into programming, enrolling in a full-time web development program and returning to DigiCo as volunteer teachers. Similarly, at INCO, programs with wraparound support saw a 44% completion rate – more than double that of programs without this support.
2. Balance upskilling with growth mindset
The world of AI is changing fast. Teaching a specific skill isn’t enough; we also have to encourage a “growth mindset” so people can adapt throughout their careers. We’ve seen this in action: TSL (The Finnish Workers’ Educational Association) found that 69% of employees in its SkillPlus program continued developing digital skills after the program ended.
Mentorship and peer networks are key: at Czechitas, alumni make up 40% of the entire educator community, creating a self-sustaining cycle of support that outlasts the initial grant.
3. Build for the long term
AI training is only as effective as the infrastructure supporting it. By providing flexible funding to local non-profits, we help them build strong, resilient programs that last. This way, we ensure the benefits of this training endure even as technology shifts.
