AI and Democracy: Unpacking the political power of algorithms

Welcome to the inaugural edition of The Hidden Politics of AI, where we explore AI’s underexplored angles and its political force shaping societies. In this first issue, we reflect on Lawrence Lessig’s recent lecture at the London School of Economics (LSE), where he unpacked AI’s effects on democracy, focusing on the 2024 U.S. election and social media’s role in polarization.

Lawrence Lessig is a prominent American legal scholar, professor at Harvard Law School, and an influential voice on issues like digital rights, corruption, and the impact of technology on democracy. He is also the founder of Creative Commons, an organization promoting open access to creative works, and a key advocate for campaign finance reform through his initiative Equal Citizens.

My Personal Motivation

This newsletter is deeply personal, rooted in my work at the intersection of gender, human rights, and technology. As co-founder of Marialab, I’ve integrated feminism and technology, creating inclusive digital spaces for marginalized groups. At Newa, I’ve led DEI strategies for over 50 organizations, improving workplace diversity and psychological safety. With a Bachelor’s in Sociology and Social Communication and a Master’s in Gender Studies from the London School of Economics and Political Science, I’ve developed a strong foundation in understanding systemic inequalities, particularly in technology.

The Political Power of AI and Engagement Models

In his lecture, Lessig warns that AI-driven business models—particularly those used by social media platforms—prioritize engagement over truth, fostering a dangerous environment for democracy. Platforms like Facebook are designed to optimize user interaction, often at the cost of exacerbating polarization. By feeding users increasingly extreme and emotionally charged content, AI algorithms drive higher engagement, leading to a polarized and angry public. This process turns public discourse into a battleground for misinformation, undermining the foundations of democratic debate.

Lessig explains that AI’s role in polarization is not always the result of malicious intent, but rather a by-product of how these algorithms are designed to maximize attention. He points out that foreign actors, such as Russian disinformation campaigns, have taken advantage of these engagement models by spreading fake news to sow discord. An example Lessig mentions is a Russian news site falsely claiming Kamala Harris was involved in a hit-and-run accident, amplifying divisive narratives. Domestic actors have also used AI to spread misinformation, as seen when Donald Trump posted AI-generated images of Taylor Swift, suggesting her support for him.

Lessig notes that AI’s influence is often twofold: one effect is intentional (from the deployers of AI systems), like targeted political campaigns that aim to manipulate public opinion, while the other is an unintended by-product of how AI engagement models operate. These models, designed to optimize time spent on platforms, often end up amplifying the most polarizing and extreme content, creating echo chambers and deepening societal rifts.

Citizens' Assemblies: A Democratic Innovation

As a solution to the threats posed by AI, Lessig emphasizes the role of citizens' assemblies, which are random, representative, informed, and deliberative groups. These assemblies allow for democratic decisions to be made without the distortions of AI-driven platforms. He provides examples of successful implementations in countries like Iceland, where citizens assembled to craft a new constitution after the 2008 financial crisis, Ireland, where they deliberated on issues such as abortion and same-sex marriage, and France, where 150 citizens convened to discuss climate policy.

Lessig views citizens' assemblies as a vital counterbalance to the distorting influence of engagement-driven media models. These assemblies are free from the pressures of social media algorithms and offer a space for thoughtful, democratic deliberation that complements the work of elected representatives. By involving ordinary citizens in key decision-making processes, citizens’ assemblies bring transparency and a direct form of democratic engagement to complex social and political issues, shielding democracy from the toxic effects of media and corporate influence.

Why This Matters

AI is not merely a technical tool—it has evolved into a political force with the power to shape societies, influence elections, and fuel division. While citizens' assemblies offer hope for deliberative democracy, we must confront the hidden politics of AI, particularly how algorithms exploit human weaknesses and drive polarization. Lessig argues that the real danger is not in AI's technological strength, but in our vulnerability to manipulative, divisive content.

The question becomes: Are the benefits of AI worth the costs to democracy? As AI systems continue to shape how information is disseminated and consumed, Lessig calls for a rethinking of these business models to ensure the survival of democracy. Citizens' assemblies offer one promising avenue to mitigate the harmful effects of AI, but broader action is needed to address the underlying engagement models that prioritize profit over democratic health.

Looking ahead

In upcoming newsletters, we’ll dive into topics such as:

• AI and sustainability: examining the environmental costs of AI training.
• Algorithmic bias: uncovering how AI perpetuates inequality.
• The political economy of data: investigating how AI is built on "private soil" and its implications for national sovereignty.
• AI ethics in chatbots: exploring the manipulative potential of anthropomorphized AI.
• The role of red teams in mitigating AI risks and influencing regulation, especially in diverse linguistic contexts.

I've spoken about AI's ethical implications through my contributions at events like Web Summit and Hacktown. This newsletter continues these discussions, inviting readers to think critically about AI’s hidden dimensions and take action toward creating transparent, ethical AI systems.

Join me on this journey to uncover the hidden politics of AI and foster a more inclusive digital society. Watch the video here for more on Lawrence Lessig's lecture.