I think public sentiment about surveillance is complicated.

People have a lot of fears about being done injury by rogue antisocial actors. Ranging from nuisances like dog shit everywhere, stolen Amazon packages and car break-ins to life-defining traumas like sexual assault. Broader surveillance, in some form, is the first step towards getting traction on these kinds of problems. Leftist commentators prefer the more "grass-roots" varieties like cell phone videos and word-of-mouth name-and-shame campaigns, but, in general, disseminating awareness of antisocial behavior and coordinating a punitive response of some kind is the goal.

I think this internal tension in public opinion pre-dates LLMs, and my personal opinion is that the surveillance applications of AI are secondary drivers of the bipartisan anti-AI sentiment, which is more about revulsion at the possibility of a future where not only "artist" but "mathematician" may be lost as vocations.