I don't think a "kill file" was the same as "moderation".
A kill file is "User X is tired of seeing User Y, so X stops reading posts from Y". Everyone else can still read Y.
A moderator is "Moderator is tired of seeing User Y, so Y cannot post to Newsgroup anymore". A much stronger action.
---------
Its been a few decades since I was part of any Usenet though. So maybe I'm mis-remembering the lingo. Or are you saying that the moderators used kill-files that somehow applied to the whole newsgroup?
Because... the later would make sense. But I really don't know how things worked back then.
There's a third layer: the servers. Newsgroups relied on servers propagating messages to other servers as well as delivering them to end users, and they could have various types of filters.
The same sort of model exists in modern distributed systems like Mastodon/ActivityPub. I can block a user from another server or an entire server individually, or my server admin can block them from communicating with anyone on my server.
What's missing that I think will be required for anything that gets popular enough is a means of sharing blocklists automatically, preferably with some machine-readable details so that they're useful even when servers have different rules (e.g. I want to subscribe to bans from Foo if the ban is for hate speech, but not for porn).
The machinery for a moderated newsgroup works by the news server you post an article to having a list of moderator email addresses for moderated groups. Instead of posting your article the server emails it to the moderator. The mod will then either just discard it, or else post it to their own news server with an Approved header to say it's moderated (and these days some crypto signing stuff). Then the post gets propagated as usual.
The "approve/deny/edit" stuff is handled by moderation software, which in principle can do anything you like.
[1] https://en.wikipedia.org/wiki/Kill_file