by Mónica Humeres, Dusan Cotoras, Renato Moretti, Iñaki Oyarzún-Merino, Teresa Correa, and Claudia López
Humeres, M., Cotoras, D., Moretti, R., Oyarzún-Merino, I., Correa, T., & López, C. (2025). Cooling down AI regulation controversies: Three closure processes in the Chilean legislative arena. Big Data & Society, 12(1). https://doi.org/10.1177/20539517241311067 (Original work published 2025)
As artificial Intelligence (AI) continues to shape critical aspects of social life, legislative debates in the Global South provide a compelling lens through which to understand what is at stake from a regulatory perspective. Interestingly, despite extensive academic debates on the inherently controversial nature of AI, public discussions often appear more consensual, with potentially contentious aspects systematically softened or sidelined.
In our article, "Cooling-Down AI Regulation Controversies: Three Closure Processes in the Chilean Legislative Arena" we analyzed 51 sessions of discussion around AI legislation in Chile, held between 2023 and 2024. Our findings reveal that socially and environmentally sensitive issues related to AI development and use are often sidelined in favor of fast-tracked consensus, in our words, cooled down. The concept is a deliberate dialogue with Dandurand, McKelvey, and Roberge’s (2024) article—published in this very journal—which introduces the notion of freezing out to describe how legacy media in Canada produce "cold controversies" around AI. While our analysis shifts the lens to legislative processes, taken together, both studies shed light on how distinct institutional settings—media and policy—contribute to the broader depoliticization of AI. In what follows, we briefly walk through the three mechanisms that illustrate how this cooling down takes shape in the Chilean legislative arena:
Deflection of Technology Liability: AI is often framed as a neutral, monolithic, or fully-formed tool—neither good nor bad—effectively disconnecting responsibility from the designers and implementers of these technologies. This perspective obscures the complex implications of the diverse applications that AI enables.
Instrumentalization of Technology Policy: Legislative priorities tend to favor agendas that relate to immediate political concerns, such as public safety while neglecting structural issues like technological sovereignty, environmental impacts, and regional consequences.
Moralization of Technology Use: Ethical principles are often used as a substitute for formal state regulation. Promoted primarily by industry representatives, this approach shifts attention away from regulatory challenges and reinforces the euphemism of “self-regulation,” favoring corporate interests over societal concerns.
Our study also highlights the critical role of expert composition in these discussions. The dual affiliations of many technical experts often introduce corporate interests, while non-computer science experts are less validated in these arenas. Furthermore, participants tend to share relatively homogeneous perspectives, limiting epistemic and social diversity.
We argue that achieving a balanced disciplinary composition and fostering plural viewpoints are essential for developing more inclusive and context-sensitive legislation. Ultimately, this article not only provides a nuanced analysis of legislative processes in Chile but also raises crucial questions about how to structure democratic debates around emerging technologies attending to contexts that are -always- socially situated.
