Answers are grounded in this paper and cite their sources.
Purpose This exploratory study examines how two conversation-based AI models, ChatGPT (USA) and DeepSeek (China), frame fundamental political concepts, “freedom, equality, democracy, social order” and the role of the state within political communication. It aims to compare the normative priorities and ideological assumptions in their responses and assess their possible relationship with the geopolitical contexts of their development. Design/methodology/approach The research used a structured comparative design and qualitative content analysis. The dataset comprises English responses to six core questions posed to GPT-4 and DeepSeek Chat in April 2024, followed by standardized follow-up questions for five items. Analysis focused on discursive framing, value priorities, and implicit assumptions. Findings Both models generally use similar democratic discourses when explaining freedom, equality, and democracy, but differ in value prioritization. ChatGPT foregrounds individual autonomy, personal rights, and democratic processes, whereas DeepSeek emphasizes collective welfare, social harmony, and the state's responsibility to maintain social order. Both refer to democracy, the rule of law, and human rights at a general level. These patterns are exploratory and not generalizable. Practical implications The study highlights the importance for policymakers, technology developers, and media professionals of understanding ideological asymmetries in AI-mediated political communication, and stresses transparency and critical awareness in conversational AI systems. Social implications The implicit normative assumptions of conversational AI can influence political information acquisition, democratic discourse, and cross-cultural understanding. Their spread in everyday information environments may shape how users interpret political concepts in unseen ways. Originality/value The study treats generative AI not merely as a technological tool but as systems situated in different geopolitical contexts that can participate in shaping social-political discourse. Comparing ChatGPT and DeepSeek offers an empirical framework for assessing ideological positioning and normative differences in conversational AI.
No takes yet. Share an insight, caveat, or question.
Emine KILIÇASLAN (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: