Whether the Lambert–Amery glacial system (LAGS) will remain near mass balance this century while adjusting to geometric change remains an open question. We couple the ice-flow model Úa to the PICO sub-shelf melt parameterization for 2020–2100 under ten high-emissions scenarios, two control experiments and targeted pinning-point perturbations. By 2100, unperturbed absolute trajectories across the twelve applied forcings contribute to mm to sea level relative to the common 2020-relaxed state; member-specific transient-minus-2020-constant diagnostics isolate the response to post-2020 forcing evolution, which spans to mm. With SMB held time-invariant over the model domain, using time-evolving rather than time-invariant 2020 basin-mean ocean conditions increases the 2100 contribution by mm. Minimizing the dynamic buttressing contribution of all pinning points adds an almost forcing-invariant mm. Across ten single-point experiments, the 2100 sea-level response is proportional to grounding-area loss. Paired perturbed–unperturbed simulations have nearly identical shelf-integrated basal melt but distinct sea-level contributions, showing that the additional response is driven primarily by reduced buttressing rather than melt-forcing changes. Overall, LAGS remains close to mass balance under the applied forcing protocol, while time-evolving ocean forcing and pinning-point weakening modulate the magnitude of the 2100 sea-level response. Dynamically important pinning points are priorities for long-term monitoring.