Political Manipulation in the Age of Interactive AI: The Case for Protecting Freedom of Thought under the ECHR

Interactive AI systems - large language models, conversational agents, and chatbots - are rapidly becoming primary interfaces through which citizens access and evaluate political information. Developed and governed largely by private corporations, these systems do not merely distribute content: they shape opinion formation through sustained, adaptive, and personalised dialogue, relocating political influence from the public sphere to the internal processes of individual deliberation. Unlike traditional forms of persuasion, such influence operates cumulatively and often opaquely within environments that simulate neutrality while asymmetrically structuring reasoning and attention. This transformation presents a challenge that European digital regulation, oriented toward systemic risk management and platform accountability, might be ill-equipped to fully capture. This chapter argues that freedom of thought under Article 9 of the European Convention on Human Rights provides an appropriate normative lens for assessing such cognitive interference. Absolute in principle yet confined in practice to settled religious and philosophical beliefs, Article 9 has left the cognitive conditions of belief formation largely unprotected - precisely the space that interactive AI now engages. The chapter advances two connected claims: first, that Article 9 should be interpreted as protecting the integrity of the processes through which beliefs and political judgments emerge, not only beliefs once formed; and second, that freedom of opinion under Article 10 offers a complementary safeguard for the transitional space between internal deliberation and external expression. Read together, these provisions establish a constitutional framework within which both private influence and state regulatory responses to AI-enabled political manipulation can be evaluated and constrained.