Background Non-communicable diseases (NCDs) kill 41 million people a year and the products and services of unhealthy commodity industries (UCIs) such as tobacco, alcohol, ultra-processed foods and beverages, and gambling are responsible for much of this preventable burden. While effective public health policies are available to address this, UCIs have consistently engaged in efforts to prevent such action across sub-national, national, and supranational levels of governance through what is known as corporate political activity (CPA). We aimed to contribute to the study of CPA and the development of effective counter-measures by formulating a set of cross-industry, evidence-informed taxonomies and a model of UCI political activity.

Methods We used five complementary methods: critical interpretive synthesis of the conceptual CPA literature; brief interviews; expert co-author knowledge; stakeholder workshops; and testing exercises using empirical literature.

Results We found 11 original conceptualisations of CPA. Four were used by other researchers as reported in 24 additional review papers. Combining an interpretive synthesis and feedback from users, we developed two taxonomies – one on framing strategies and one on action-based strategies. The former identified three frames (policy actors, problem, solutions) and the latter six strategies (access and influence in policymaking, use the law, manufacture support for industry, shape evidence to manufacture doubt, displace and usurp public health, manage reputations to industry’s advantage). We also offer an analysis of the strengths and weaknesses of industry strategies and a model that situates industry CPA in the wider social, political, and economic context.

Conclusion Our work confirms the similarity of CPA across UCIs and captures its extensive and multi-faceted nature, the disproportionate power of corporations in policy spaces and the unacceptable conflicts of interest that characterise their engagement with policymaking. We suggest that industry CPA is recognised as a corruption of democracy, not an element of participatory democracy. Our taxonomies and model provide a starting point for monitoring, further research, and the development of effective solutions.