Beyond the Deepfake Panic
For years, the conversation around AI in elections has been dominated by the threat of 'deepfakes'—hyper-realistic but fake videos or audio clips of politicians. This concern is valid; AI-generated misinformation can easily spread, confuse voters, and
erode public trust. However, fixating on fake content alone misses the bigger picture. The more profound impact of AI on our democratic processes is quieter, more structural, and arguably more potent. It lies in the technology's ability to create and amplify power imbalances, fundamentally altering who holds influence and whose voice gets heard.
The New Oil: Data and Asymmetry
In the digital age, data is power. Political campaigns have long sought to understand voters, but AI supercharges this effort. Parties and candidates with access to vast datasets—gleaned from social media activity, consumer habits, and data brokers—can use AI to analyze and predict voter behavior with unprecedented accuracy. This creates a massive asymmetry. Well-funded political machines can build sophisticated profiles of individual voters, understanding their fears, hopes, and triggers. In contrast, smaller parties, new candidates, or grassroots movements without access to this data infrastructure are left at a significant disadvantage, struggling to make their message heard above the noise. The result is a political landscape where power gravitates toward those who can afford the best data and the most advanced AI tools to interpret it.
Hyper-Targeting and Political Bubbles
This data advantage fuels the engine of microtargeting. AI allows for the creation of thousands of variations of a political message, each tailored to a specific demographic or even an individual. While one voter might see an ad focusing on economic concerns, their neighbour might be served a completely different message about social issues, all based on their personal data profiles. This practice, far more precise than traditional advertising, can be highly effective at persuading voters. However, it also has a corrosive effect on democracy. It fosters political 'filter bubbles', where voters are only exposed to information that reinforces their existing beliefs, increasing polarisation. It also makes it difficult to hold politicians accountable, as they may be sending contradictory messages to different groups of voters simultaneously, all without broad public scrutiny.
The Widening Resource Gap
Sophisticated AI tools are not cheap. Developing, training, and deploying advanced AI models for political campaigning requires significant financial investment and technical expertise. This reality widens the gap between the 'haves' and the 'have-nots' in the political arena. Wealthy campaigns and established parties can afford to hire data scientists and purchase access to premium AI services, allowing them to optimize resource allocation, refine messaging, and increase voter engagement. This can give them a decisive edge. Conversely, less-resourced campaigns may be unable to compete, not because their ideas are less compelling, but because they lack the technological firepower to reach voters effectively. This creates an environment where financial power increasingly translates into political power, challenging the principle of a level playing field in elections.
Algorithmic Bias in Governance
The impact of AI power imbalances extends beyond campaigning and into governance itself. As governments begin to adopt AI for public services, resource allocation, and policy analysis, the risk of algorithmic bias becomes a major concern. AI systems are trained on data, and if that data reflects existing societal biases, the AI will learn and perpetuate them. For example, an AI model used to allocate public resources could inadvertently discriminate against certain communities if its training data is skewed. This means that decisions affecting people's lives could be made based on biased algorithms, further marginalizing already disadvantaged groups and concentrating power and resources in the hands of the few. Ensuring fairness and transparency in public-sector AI is crucial to prevent the technology from reinforcing the very inequalities governments are meant to address.
