September 15 was the International Day of Democracy, and one of the promises of artificial intelligence is to help fix our broken political systems. But are we solving the right problems? Is AI solving them? As part of our continued work to place justice at the heart of innovation, Ronald Lenz, HiiL’s Director of Innovation, and Martin Gramatikov, Director of Knowledge and Research, reflect on good governance through AI.
Something strange happened in Albania last week. The Prime Minister announced a new cabinet appointment: an AI bot named Diella, tasked with combating government corruption. No personal agenda, no bribes, no scandal-prone expense accounts. Just pure, algorithmic honesty.
It sounds almost too good to be true. And maybe it is.
We’re living through a moment when our democratic institutions feel painfully outdated. While we can order dinner, find a date, and stream movies with a few taps, participating in democratic processes still feels like navigating a Byzantine maze. Voting requires research most of us don’t have time for. Exercising rights and obtaining justice means hiring lawyers most of us can’t afford. Making your voice heard in government often requires knowledge, resources and connections most of us don’t have.
So when Silicon Valley promises to streamline all of this with AI, it’s tempting to say yes.
Justice in Your Pocket
Picture this: You’re fighting a bogus credit card charge that’s been driving you crazy for weeks. Instead of spending hours on hold with customer service, you open an app called « YourLegalAgent » and let an AI handle it. The bot reads your case, communicates with the bank’s systems in perfect legalese, and gets you a refund in minutes. This isn’t science fiction, startups are already building systems like this. And honestly? It sounds pretty great. For many, the justice system appears too slow and inaccessible, almost nonexistent. It’s too expensive, too slow, and too complicated to navigate without costly professional help.
But here’s the catch: What happens when your free, ad-supported AI lawyer goes up against the bank’s premium, corporate-grade AI? Is that really justice, or just a fancier version of the same old inequality?
And there’s something else that bothers us about algorithmic legal information and advice. Real justice isn’t just about efficiency and automatic application of rules to facts; it’s about understanding context, respecting social and communal values, and showing mercy while recognising that every dispute and its parties are different. When we turn conflict resolution into a battle between competing algorithms, do we lose something essentially human about how we solve problems fairly?
The Robot Politician
The stakes get much higher when this logic moves from the use of services to governance. Albania’s AI minister might eliminate corruption in government contracts, which would be genuinely wonderful. But imagine if that technology went further. Elections are a time for tough individual and collective decisions. What if there was an app that could analyse every candidate, every ballot measure, every policy position based on your personal values, then tell you exactly how to vote? Some entrepreneurs are already working on versions of this. The pitch is seductive: Why spend hours researching candidates when an AI can do it for you?
But democracy isn’t supposed to be convenient. The messy work of reading up on candidates, arguing with your neighbours, changing your mind—that’s not a bug in the system. That’s an intrinsic part of the game. Similar to justice, in the exercise of democracy, the process matters no less than the outcome. It’s how we learn to live together as citizens, rather than just consumers.
If we outsource our political thinking to machines, we risk becoming spectators in our own democracy.
The Choice Is Still Ours
None of this means we should reject AI in civic life entirely. Technology genuinely helps solve real problems, making justice and government services more accessible, reducing corruption, and helping people navigate complex systems. The question is why and how we do it. We need to ask ourselves: What are our individual and collective values? How can AI assist us in putting these values into practice? Do we want civic AI systems we can actually see and understand, or are we comfortable with our democracy running on secret algorithms? Do we want these tools to be built by a handful of tech giants, or should communities have more control over the systems that govern them?
Most importantly: Are we trying to make better citizens or just more efficient ones?
The technology isn’t going to answer these questions for us. Despite all the hype about AI, it can’t solve the fundamental challenge of democracy, figuring out how to live together despite our differences. That’s still up to us and will always be.
