Let’s face it: artificial intelligence has quietly woven itself into nearly every corner of our daily routine. It’s the invisible hand behind the videos we binge, the ads we scroll past, and the headlines that grab our attention. So it’s only natural to wonder if these algorithms are so good at shaping what we buy and watch, whether they can also shape who we vote for.
That question is more urgent than ever. Elections rest on a simple democratic promise: we get to choose for ourselves. And yet, those choices are increasingly being made inside a digital ecosystem where algorithms, social platforms, and AI are pulling the strings behind the curtain.
AI is already on the campaign trail:
No one is claiming that a chatbot physically steps into a voting booth and ticks a box. Its influence is far sneakier.
Social media platforms decide which posts, clips, and political pitches land on your feed. During an election cycle, that means they help decide what you see and don’t see about candidates and the issues that matter. With generative AI in the mix, the game has changed even further. A polished campaign video, a batch of persuasive memes, a tailored message for every kind of voter—all of it can be churned out in seconds.
As International IDEA points out, this technology can be twisted to distort facts, fabricate deepfakes, and micro-target communities with highly specific political messaging.
Picture two neighbours living on the same street. One is constantly fed stories about immigration; the other gets an endless stream of videos about jobs and climate. They’re participating in the very same election, yet they might as well be living in different political universes. Neither has been told outright whom to vote for—but their sense of what actually matters has already been shaped.
The deepfake dilemma:
Perhaps the scariest development is the rise of deepfakes. With today’s generative tools, you can produce alarmingly realistic footage of a politician saying or doing something they’ve never done. And once that content hits social media, it can spread like wildfire.
The real problem isn’t just that people might be fooled. It’s that when anything can be faked, people start doubting the real stuff too. UNESCO and UNDP have flagged AI-generated disinformation, privacy violations, hate speech, and deepfakes aimed at political figures as major threats to elections.
Democracy runs on trust. Voters need to believe that the information they’re getting is basically sound, that the process is fair, and that the final count reflects real choices. If AI makes it harder every day to tell fact from fiction, that trust so essential to everything starts to crumble.
But hold on: AI isn’t the villain here:
It would be far too easy to paint AI as pure danger. That’s not the whole story.
AI can also be a genuine ally to electoral institutions. It can sift through massive datasets, flag unusual patterns, make election information easier to find, and streamline administrative work. International IDEA puts it simply: AI has a dual nature. Without proper safeguards, it can disrupt democracy; with the right oversight, it can strengthen it.
Imagine AI helping citizens navigate voting rules, translating election materials into dozens of languages, or making information accessible to people with disabilities. That’s a huge win for inclusion.
So the real question isn’t whether we should ban AI from elections entirely—that ship has sailed. It’s how we set boundaries, and who gets to decide where those boundaries lie.
Who’s accountable when an algorithm sways voters?
And that brings us to the thorny issue of responsibility. If a political party uses AI to spin misleading content, who takes the blame? The candidate? The party? The platform? The tech company that built the tool?
Regulators are starting to catch up. In the European Union, new rules on political advertising came into force on October 10, 2025. Political ads now have to say who paid for them and whether they’re using targeting tricks. The rules also tighten the reins on how personal data can be used for political marketing.
Why does this matter? Because transparency gives voters something crucial: context. When you know who’s trying to nudge you, you’re far better equipped to see through the nudge.
Of course, laws alone can’t fix everything. Technology moves at the speed of light, while legislation crawls at the speed of bureaucracy. And a manipulated clip can cross borders in a heartbeat.
So, can AI really decide our leaders?
Not directly, at least not in the old-school sense of “deciding.”
Real people still cast the votes. But in between, there’s an ever-growing digital machinery—recommendation engines, hyper-targeted ads, social networks, generative AI, and oceans of personal data. These systems shape which issues get seen, which messages go viral, and which information reaches each voter.
That means AI’s real power in politics isn’t about replacing voters. It’s about shaping the very world voters make their decisions in.
And that distinction makes all the difference.
The future of democratic elections hinges on one thing: keeping humans firmly in charge of political decisions, while laying down clear, enforceable rules for algorithms and AI-generated content. That job doesn’t belong to any single group; it’s a shared responsibility for governments, tech companies, journalists, election officials, and citizens alike.
AI can be a useful tool for democracy. But it should never, ever become the quiet force deciding our leaders.
At the end of the day, elections ought to stay a human decision. We can let technology inform that choice—but we should never let it take the wheel.
Sources:
- UNESCO – Artificial Intelligence, Freedom of Expression and Elections
http://www.unesco.org/en/articles/new-unesco-undp-issue-brief-highlights-impacts-ai-freedom-expression-and-elections - International IDEA – Artificial Intelligence for Electoral Management
http://www.idea.int/publications/catalogue/artificial-intelligence-electoral-management - International IDEA – Protecting Democratic Elections through Safeguarding Information Integrity
http://www.idea.int/publications/catalogue/protecting-democratic-elections-through-safeguarding-information-integrity - European Commission – Transparency and Targeting of Political Advertising
http://commission.europa.eu/strategy-and-policy/policies/justice-and-fundamental-rights/democracy-eu-citizenship-anti-corruption/democracy-and-electoral-rights/transparency-and-targeting-political-advertising_en - UNESCO – Artificial Intelligence and Democracy
http://www.unesco.org/en/articles/artificial-intelligence-and-democracy?hub=83294
