Building Skilled Governments: Unlocking AI's Promise for People Worldwide

A conversation between Maggie Johnson, Global Head & VP, Google.org and Robyn Scott, Co-Founder & CEO of Apolitical, exploring how governments worldwide can support their teams to build new skills to fully unlock the AI opportunity to benefit the communities they serve.

Apolitical has a unique vantage point as it engages with governments around the world: How are public servants and leaders reacting to AI right now?

RS: AI is changing almost everything, and most governments just aren't ready. The potential consequence if governments fall behind, is a tremendous—generational—lost opportunity for prosperity.

Apolitical’s AI Campus is a very ambitious initiative to quickly close that gap and to build AI-ready governments. We want to go after these enormous opportunities to deliver better services for citizens in a world where every government is facing squeezed public budgets. Our goal is to support a million public servants globally to be skillful with AI. We're enormously grateful to be supported by Google.org in building it across all parts of the world, from Kenya to the UAE.

MJ: We’re proud to do it – it’s incredibly critical work, especially given the pace of progress with AI and the enormous potential benefits for communities worldwide.

I taught computer science for many years, and one of the things that was always a big part of supporting new computer scientists was helping them understand how important it is for them to not only be practitioners in the field, but also to help others, especially policymakers, understand the technology.

RS: Absolutely. We think a lot about how to close the gap, and we’re proud of the university partnerships we’ve been able to build as part of the AI Campus strategy. In this way, public servants aren't necessarily always on the back foot; they are seeing what's coming, and getting some support.

And I'll just stress one thing because it is vivid in my daily and weekly conversations with leaders and governments around the world: leaders are very behind.

On the one hand, you've actually got quite a positive attitude in general in government to AI. Public servants are citizens like all of us, and they are just as excited as all of us are about how AI could solve some of the bottlenecks we have in public services delivery. And leaders' rhetoric is very positive towards AI.

However, when you actually talk to them, leaders in government are quite unsettled by not understanding it. They don't feel they have safe spaces to ask what they think might be ‘dumb’ questions.

We need to create those spaces and ensure leaders understand just the colossal pace of change. And it’s equally important for leaders to engage with the technology on a daily or weekly basis to feel it and to understand the rate of change.

As you mention, the technology is changing quickly, and the public sector has such a diversity of roles - from an education and content perspective, how do you keep up? 

RS: We’ve developed a broad curriculum, but we also focus on networks and connectivity and communities of practice: the greatest insights often lie with the practitioners—what's being attempted, what are the bottlenecks.

And when it comes to AI, the technology is moving so quickly that you are really only as smart as your network. So building live, organic communities that support discovery, support rapid knowledge sharing between public servants is a key part of the solution.

An underappreciated dynamic with public servants is that there aren't competitive barriers to sharing in government, for the most part. People join government with an immense sense of purpose. One of the single most powerful rewards is knowing that something you've done is being scaled and adopted elsewhere, particularly in another country. We lean into that motivation.

What are you seeing in developing countries in how they approach AI?

MJ: I imagine there are greater distinctions between developed and developing countries. You mentioned you're operating globally, so can you talk a little about what you're learning in emerging economies?

RS: The opportunity for AI in developing countries is enormous, and there is an eagerness for it. Public servants in emerging markets are the most enthusiastic, because they look at the fragility of their health and education systems, and they see AI as a path to offering radically different public services.

I think there's an incredible moment where Africa can repeat what it did with mobile phone technologies – making a virtue out of absence to achieve phenomenal leapfrogging.

When I was in Addis, I went to Ethiopia’s AI institute there; founded in 2020, and it's astonishing to see the specific applications that they've built. One tool that stuck with me particularly is around breast cancer detection. American and Asian models weren't working for African women, so they shared with me how they built their own lightweight models and screened 5 million women in six months, with 98% accuracy. It's completely transformative.

For governments thinking about how to help their teams become AI-ready and AI first, what should be top of mind? 

MJ: Your recent Gen AI survey showed large gaps between public servants using AI versus those actually trained to use it—something like 59% using it versus 15% trained. Is that a problem, and how do you think about alleviating that knowledge gap?

RS: It's a massive problem, but I don't want to throw the baby out with the bathwater on this. It's a great thing that so many people are using it organically. Experimenting with using AI responsibly is the first step of the journey.

But, the training gap is a huge one. New data from a large Apolitical survey of more than 4,000 public servants around the world showed that of those who were implementers in government—the people rolling out AI—only 35% had familiarity with safety frameworks.

That’s a very scary number, and we need to close that gap. A lot of time is spent at the regulatory and political levels, and even once laws get passed, you don't get the cascade of guidance down, so institutions are none the wiser. Making complex guidance more accessible is actually a crucial part of training.

Looking ahead - what does the future hold for AI-first public service?

RS: A recent shift we’re seeing from the past few months is the difficulty of going from pilots to scale. We’re seeing that governments aren't quite sure where to place their bets. This "pilotitis" is a classic problem, and a big area of need is helping governments see who's succeeding and failing globally to help pick their bets and decide what solutions or efforts to scale.

MJ: And there’s so much good that can come from those bets and solutions. At Google, we’re really enthusiastic about the potential of AI to help governments work more effectively across everything they do – whether it’s providing essential daily services, tackling tough issues like extreme weather events, boosting economic growth, or keeping communities safe. Thinking about what’s possible - what do you hope the future looks like two years from now?

RS: I think what's happening now is similar to the era of digital transformation a couple of decades ago. We are now at a point where we're entering the era of AI and “AI-ready” governments. Success for us looks like changing a whole system, by helping entire workforces to learn new skills and to be able to meet this opportunity.

To make it concrete: supporting some stories that will be on the front page of newspapers in their countries because something genuinely transformative has happened for citizens at a transformative scale.

In one of our recent surveys, we asked public servants—a global sample—"How much do you agree with this statement: 'My government is moving fast enough to meet citizens' needs'?" Two-thirds of them disagreed.

But I think what that means for government is that if you can help people to be better, to have more impact, to unblock bottlenecks, you can be assured that there's an enormous appetite amongst public servants for that - and that’s what AI does.