You are assuming that harassment was direct, sending messages to the accusers. My impression is that this was a group chatting privately about the accusers either making fun of them or coordinating actions.
While this can be done through other channels (in person or private cellphone) allowing it on corporate infrastructure without monitoring is not acceptable.
If harassers, as you have mentioned, can simply switch to another channel, then what problem exactly is this measure trying to solve? I'm honestly confused.
Before this change, would the company have been liable for harassing DMs on Slack, an unaffiliated service where they previously were not able to read peoples DMs?
Friction. Yes they could use PGP messages. But most people are idiots and use the systems that are available and simple. If somebody needs to think "I've got to go through these extra steps to be an asshole" then they might think twice in the first place.
Now the harassers are doing more and more to harass and leaving more evidence behind. And you have a more compelling case, "they were on Slack, they were on email, they were texting, they set up their own ICQ server..."
What's stopping the harassers from setting up a secondary Slack install for coordination of such nefarious ends?
The only use I can see for this would be evidence after the fact. Surveillance is almost never the proper way to enforce acceptable standards of behavior.
> What's stopping the harassers from setting up a secondary Slack install for coordination of such nefarious ends?
Company policy for one could be applicable.
You’re missing the point though, what you do with non-company provided tools is held to a different standard from officially blessed and sanctioned ones.
While this can be done through other channels (in person or private cellphone) allowing it on corporate infrastructure without monitoring is not acceptable.