“…Natural observations of this process, however, are notoriously rare. Despite numerous intra‐oceanic “hot‐spots” (Courtillot et al., 2003), only one of them has triggered subduction around the Caribbean plate in the Late Cretaceous (Whattam, 2018; Whattam & Stern, 2015), whereas in all other cases no proven hot spot‐related ocean‐ocean subduction has been detected so far. One of the most prominent “negative” examples is the North Atlantic Igneous Province (Saunders et al., 1997) associated with the Iceland plume that 1) produces melt during most of the past 60 Myr (e.g., Kerr, 2003), thus contributing to melt‐induced magmatic weakening of the overlying plate, which is typically the most important condition for plume‐induced subduction (Ueda et al., 2008, Gerya et al., 2015; Baes et al., 2016, 2020a, 2020b); 2) has presumably triggered the opening of the North Atlantic Ocean at ∼60 Ma (Beniest et al., 2017b) and caused the active spreading axis to jump from the Aegir ridge to the Kolbeinsey ridge at ∼30–35 Ma (Koptev et al., 2017), thus attesting a sufficiently high buoyancy of the plume to overcome the strength of not only oceanic but even continental lithosphere; 3) channeled along the thin‐lithosphere corridors underneath continental lithosphere and the mid‐oceanic ridges (Steinberger et al., 2019); 4) presently extends not only along the mid‐oceanic ridge but also in perpendicular directions (see seismic tomography by Rickers et al., 2013 and laboratory and numerical experiments by Schoonman et al., 2017 and Koptev et al., 2017, respectively), thus impinging oceanic lithosphere of ages older than 20 Ma as required for the formation of stable subduction zones in numerical experiments (Baes et al., 2016; Lu et al., 2015).…”