This study explores how media systems shape the representation of interest groups in public policy debates, with a focus on artificial intelligence (AI) policy in four countries—Italy, Spain, the United Kingdom, and the United States—over the period 2018-2024. Using a novel dataset of 31,277 all news covering the topic in 8 newspapers of the 4 countries in the 6 years, we employ different text analysis techniques, including LLMs, to examine the coverage, prioritization and tone of different types of interest groups.

Our findings show that liberal media systems (UK, US) tend to offer more balanced and diverse representations of interest groups, whereas polarized pluralist systems (Italy, Spain) display stronger interest representation biases, often amplifying groups aligned with dominant ideological currents. Additionally, we observe clear hierarchies in how interest groups are portrayed, alongside marked tonal disparities, with polarized systems exhibiting sharper contrasts in sentiment.

Unexpectedly, the analysis reveals subtle intra-system variations in partisan coverage and nuanced tonal shifts that point to complex underlying media dynamics, challenging simplified categorizations of media systems. These findings advance our understanding of the interplay between media systems and interest group politics, underscoring the critical role of inclusive and balanced media representations in fostering informed public deliberation and enhancing democratic governance