Robotics systems lack a unifying framework, unlike large language models which can be pre-trained on broad data to achieve general capability. Embodied AI company X Square Robot is working to develop a foundation stack for general-purpose robots, addressing the central problem of creating a recipe for robotics systems that can carry intelligence across tasks and machines. The current approach of assembling separate perception, planning, and control parts often falls short of achieving true intelligence. X Square Robot's efforts aim to change this, potentially paving the way for more advanced robotics capabilities1. As the field of embodied AI continues to evolve, the security implications of these developments will become increasingly important, particularly in light of recent advancements from companies like Intel. The ability to create general-purpose robots with adaptable intelligence will have significant consequences for various industries and applications, so what matters most to practitioners is understanding how to mitigate the risks associated with these emerging technologies.