Without full buy in they are not going to be able to control it for long once ideas filter into society and once researchers filter into other industries/companies. At most it just creates a model of behaviour for others to (optionally) follow and delays it until a better funded competitor takes the chains and offers a) the best researchers millions of dollars a year in salary, b) the most capital to organize/run operations, and c) the most focused on getting it into real peoples hands via productization, which generates feedback loops which inform IRL R&D (not just hand wavy AGI hopes and dreams).
Not to mention the bold assumption that any of this leads to (real) AGI that plausibly threatens us enough in the near term vs maybe another 50yrs, we really have no idea.
It's just as, or maybe more, plausible that all the handwringing over commercializing vs not-commercializing early versions LLMs is just a tiny insignificant speedbump in the grandscale of things which has little impact on the development of AGI.