It’s obvious to those of us established in this space that “cyber research” is being used as a tool for policy arguments and regulatory capture. Your GLM-5.3 blog acknowledges that it’s a necessary recourse for cyber defenders while primarily characterizing it ‘as dangerous as Mythos’. Given the recent vilification of open weights, releasing this blog on the same day as the President’s AI summit feels like a transparent willingness to throw defenders under the bus in pursuit of an AI regulation outcome. There’s a variety of massive categorical problems that cybersecurity desperately needs to address. AI could provide the resources for that. An IoT botnet running a distributed model at sub 10-tokens per second is not representative of any realistic concern. It’s fodder to substantiate some bizarre narrative of self-replication from a sci-fi threat model and we don’t think it’s cute.