Artificial Intelligence (AI) is a contested and rapidly evolving domain shaped by discourse, power, and historical narratives. Using Foucauldian genealogy, Bacchi’s poststructural policy analysis, and Habermasian legitimacy theory, the analysis reveals how framing practices obscure ethical complexities while participatory sensemaking offers a more inclusive path forward. AI is not a fixed entity but a dynamic concept constructed through competing narratives and regulatory framings. Linguistic choices and policy discourse influence public perception, governance priorities, and the distribution of risks and benefits. The EU’s AI Act exemplifies how normative assumptions and legitimation struggles manifest in transnational regulation. Ethical design of AI systems requires more than technical solutions—it demands democratic engagement, contextual awareness, and amplification of marginalized voices. By embracing discourse as a productive and contested process, ethical sensemaking can guide the development of AI systems that are transparent, just, and responsive to societal needs.