AI safety advocates do not typically demand to be given personal political power, only for a set of preferred policies to be enacted, like anyone advocating for any political cause.
The question is who will be enacting and controlling those policies. Somebody like METR? That's the people. And yes, every regulatory body has these dangers - but the amount of control demanded for this body is quite unusual - it's full veto right over the whole industry.
I’m trying really hard to avoid the snark.
History provides ample examples and reasons to be deeply suspicious of movements that presented themselves as merely pursuing a program or cause - but that in practice turned out to be about control of political office