The Artificial State and AI: The Impact of Algorithms on Democracy

The article analyzes the transition from traditional political persuasion to advanced algorithmic manipulation. The author distinguishes between three operations: knowing, predicting, and shaping preferences, pointing to the risk of enclosing them within a single feedback loop (the Artificial State). Unlike mass communication, microtargeting and conversational AI systems allow for adaptive persuasion that bypasses public scrutiny. Cited research from 2024 and 2025 confirms that interactions with AI models can realistically shift electoral preferences in the USA, Canada, and Poland, making the automation of persuasion on a mass scale a serious institutional problem. The text raises questions about the limits of permissible influence and the necessity of protecting the epistemic sovereignty of the individual against predictive technology.