Abstract While media coverage of climate change has been shown to imply selective knowledge transformation ( Carvalho 2007 ; Brand & Brunnengräber 2012 ; Kunelius & Roosvall 2021 ), studies assessing the potential for climate experts’ terminology to acquire ideological undertones as it enters mediatic discourses are still scarce. Through this article, we aim to compare the meaning climate experts and the media give to terms pertaining to climate change in English discourses and to determine whether potential cotextual variation in the discourses produced by these two communities have ideological implications. To this aim, we use the deep learning algorithm Word2vec ( Mikolov et al. 2013 ; González Granado 2021 ) to identify terms whose cotext of occurrence is prone to high variability depending on whether it is included in a newspaper corpus on climate change or one composed of reports from intergovernmental organizations. We then rely on statistical tools from corpus linguistics to compare the main co-occurrences of two of the terms identified – adaptation and energy security –, which we combine with Critical Discourse Analysis ( Baker et al. 2008 ) to interpret the variation in terms of meaning and ideological significance. Results suggest that the appropriation of expert terminology by the media does entail a certain degree of conceptual variation, which notably seems to allow for bringing issues of social justice, financing and energy transition into focus and assessing expert knowledge along those lines.