The first is about testing for banned words deliberately.

The second link appears to be about anti-spam suspensions related to people who controlled many different Twitter accounts simultaneously using bots, not content related enforcement.

The third link is about the following tweet:

"Can’t wait for November 4th when millions of antifa supersoldiers will behead all white parents and small business owners in the town square"

That isn't an "innocuous reply to harrassment" as you implied in your first comment.

So far I don't think you've demonstrated the point you were trying to make. None of the examples of these suspensions were replies to harassment.

> The first is about testing for banned words deliberately.

After being banned for using those words inocuously.

> That isn't an "innocuous reply to harrassment" as you implied in your first comment.

No, that wasn't one of those. It was clearly a joke, however.

> None of the examples of these suspensions were replies to harassment

A fair point. Next time I log into Twitter, I'll make a collection and come back with those, ok? I just didn't happen to have one ready for this discussion at current time.

Intentionally getting banned by doing something publicly known to be bannable doesn't really apply here httpswitter.com/zozie_kicks/status/1078008088978956289

Well, it was a test to see if it was the word 'goy' that was the trigger - previous suspensions had already happened. But fair point, it's not a great example.