TLDR: This research paper, “Digital Domination: A Case for Republican Liberty in Artificial Intelligence” by Matthew David Hamilton, argues that AI, particularly through digital advertising and social media algorithms, poses a significant threat to republican liberty—freedom from unaccountable power. It introduces the concept of “digital domination,” where algorithms can subconsciously influence individual behavior and thought, and multinational tech companies can sway political processes without effective accountability. The paper highlights how this arbitrary power operates at both individual and political levels, providing examples like targeted ads, social media’s emotional contagion, and foreign influence in elections. It concludes by advocating for increased user control over algorithms and the establishment of supranational institutions to regulate AI and protect this crucial form of freedom.
Artificial intelligence (AI) is rapidly transforming our social and political landscapes, prompting crucial discussions about the ethical principles that should guide its development and regulation. A recent paper, Digital Domination: A Case for Republican Liberty in Artificial Intelligence, by Matthew David Hamilton, argues that AI already poses a significant threat to a specific kind of freedom known as republican liberty.
Republican liberty is defined as freedom from unaccountable power, or ‘domination’. Unlike other concepts of freedom that focus on the absence of direct interference, republican liberty emphasizes that individuals can be unfree even if they aren’t directly obstructed, simply because a more powerful entity has the capacity to interfere with impunity and at will. Think of a slave with a benevolent master who never interferes; the slave is still unfree because the master holds arbitrary power over them. This paper highlights how AI creates a new form of this unfreedom: digital domination.
AI’s Influence on Individuals
At an individual level, AI algorithms, particularly those in digital advertising and social media, can subtly influence our behavior and thoughts. For instance, someone shopping for an engagement ring online might use an incognito browser to avoid targeted ads that could spoil the surprise. While no one is directly stopping them from shopping, their behavior is altered due to the algorithm’s potential for interference. Users often have limited power over these algorithms, which operate at the discretion of their creators, often without considering the user’s well-being.
Social media algorithms also pose an ideational threat, shaping our perceptions and emotional experiences. Research has shown how manipulating news feeds can lead to emotional contagion, and how algorithms can amplify controversial content or even radicalize users. Concerns about platforms like TikTok, owned by foreign companies, illustrate fears that content could be promoted or suppressed to influence populations without any accountability to the affected public. In these scenarios, algorithms have the capacity to interfere with our emotional experiences and political beliefs, influencing us subconsciously and without direct recourse.
AI’s Impact on Politics
Politically, AI creates domination through two main avenues: the multinational nature of technology companies and the rapid pace of innovation. Multinational tech giants like Meta (Facebook) and X (formerly Twitter) often find themselves in positions to influence domestic political conflicts in various countries. For example, the Myanmar military used Facebook to incite violence, and X censored content in Turkey before an election. In such cases, these companies’ executives make decisions that impact entire political communities, yet they are largely shielded from accountability by national polities.
Furthermore, the speed with which AI companies innovate makes it difficult for governments to keep pace with regulation. This often leads to self-regulation, as seen when OpenAI blocked an AI chatbot mayoral candidate. While such decisions might seem reasonable, they demonstrate private companies making significant political choices without public oversight. This lack of robust government oversight exposes individuals to domination, as corporations wield arbitrary power over political life.
Also Read:
- European AI Developers Grapple with Privacy: A Look at Risks and Real-World Solutions
- Navigating the AI Policy Maze: A New Framework for Computer Science Education
Protecting Republican Liberty in the Age of AI
The paper concludes by offering concrete recommendations to protect republican liberty. These include increasing individuals’ ‘antipower’ – their ability to hold algorithms and their creators accountable. This can involve giving users more control over their data and algorithmically determined recommendations, such as clearer explanations for suggestions and easier ways to override them. Apple’s “Ask App Not to Track” feature is an example of such a step.
More importantly, protecting republican liberty requires creating political institutions that can hold algorithms and their developers accountable in the real world. Given AI’s global reach, the paper advocates for supranational institutions, similar to the European Union’s Digital Markets Act and Digital Services Act, which can effectively regulate transnational and multinational actors. Without such global architecture, domestic political institutions may be largely ineffective against the most significant violations AI might perpetrate. By integrating AI with republican liberty in mind, we can strive for a future where AI contributes to human society without eroding our fundamental freedoms.
