Section 230 needs an update. It needs to be made clear that hosing speech is not a liability.
"Recommending" speech with a black box algorithm that the user can't control or select IS a liability.

I was watching the bipartisan senate hearing on 230. Even Ted Cruz, and every witness there, said 230 should exist, but it needs AMENDED! And they want it to include AI companies. Meaning they'll be responsible for the content that users create.

I did a reacts video on this for folks that ate interested:

https://tubefree.org/w/aKkLkyiz8Spf2csNfgUkdK

Liability or Deniability? Platform Power as Section 230 Turns 30

' Upon announcing the hearing, Sen. Cruz said: “Big Tech—the most powerful companies on Earth—can exercise monopoly power to make views they dislike disappear and that should scare everyone. When it comes to viewpoint suppression, however, repealing section 230 might increase censorship. I look forward to hearing from our witnesses and discussing possible reforms to section 230 so online platforms are a free and open marketplace for ideas.” '

I'm usually a big fan of the EFF, but it's wrong on this one. If you decentralize to the limit -- i.e., such that each user is running their own instance for themselves -- it becomes okay for the service to become liable for the user's speech because the user and the service owner are one in the same. In reality, (extremely) federated social media is the only kind that can survive without Section 230 and thus repealing it entirely would be a win for the Fediverse.

(You could argue "but users won't go to the trouble of running their own instance," but to that I'd say "they will if the law doesn't give them any other choice, short of not participating at all.")