The most typical way this is done is through selective enforcement. You create rules that most of anybody can be argued to be breaking, but only enforce it for individuals or organizations that you have an issue with. With net neutrality it's pretty easy to see how this would happen. The previous net neutrality rules, under Tom Wheeler, included a carveout for nearly all rules under anything that could be justified as "reasonable network management."
This carveout is necessary due to the nature of network management, but it creates an interesting scenario. That term was used literally dozens of times, but also defined absolutely nowhere. This now gives politicians the ability to interpret it, as they see fit. For a company that runs afoul of the powers that be, suddenly it's a standard that can't be met. For those in favor, anything can be accepted as "reasonable network management."
And this is how corruption and tyranny typically works works. The cases of a somebody going completely rogue, or a politician doing something after being handed golden bars or a bag of cash - do still happen. But, by and large, the system is big enough, and complex enough, that if somebody with enough power wants to screw somebody - they can do it legally. And if somebody protests inequitable or unfair treatment, just call it a whataboutism.
I am always surprised when somebody finds this a novel take. This sort of corruption is older than Boss Tweed. But I guess every generation has to relearn the same lessons.
It's all about risk vs reward for both parties. If a cop pulls you over for a speeding ticket, trying to bribe him is a high risk option that could suddenly see you facing felony charges to try to get out of a low cost ticket. And similarly for the officer. You're not going to pay more than the ticket in the bribe, certainly not much more, so it's negligible, and accepting it could completely destroy his livelihood if found out.
On occasion it still does happen, but it tends to be because the mutual risk/reward ratios are changed. One of the most common bribes cops are found taking is sexual stuff. That may have a much higher value for the officer than the bribe price would, and a much lower risk for the briber since they can be overtly flirtatious without ever making it clear (from the point of view of a criminal prosecution) they're offering to exchange some sexual favor for a way out.
By contrast in the government scenario the risk for both parties is effectively zero since much will go on without a single word even needing to be said. When various government agencies were pressuring Twitter to censor all sorts of things, they never made a single overt threat - yet they simultaneously spoke both threateningly and demandingly. It's the same sort of relationship. The people making the "requests" of Twitter had a million ways of going after Twitter, or the people in charge of it, and Twitter knew this. So they played ball. Even if you got access to every single email, you wouldn't be able to prosecute anybody over it. It was all completely legal, which creates a near infinite risk:reward ratio.
