Who in their right mind is going to blindly take the code output by a large language model and toss it on a cruise missile? Sleeper agents are trivially circumvented by even a modicum of human oversight.
The weights and data pipeline are open sourced and described explicitly in the paper they published. The non-reasoning data isn't nearly as interesting as the reasoning data though
I'm sure the powers-that-be will absolutely pay attention to that clause.