Facebook is not hiding behind Section 230 for their recommendation algorithm. They are hiding behind the First Amendment. By making recommendations, Facebook is expressing an opinion. US courts take a dim view of government attempts to restrict the expression of opinions. There is an extensive body of case law on this issue.
If Facebook's algorithm recommends illegal porn or a terrorist's call to arms (and it has done both), then the First Amendment doesn't protect it. Section 230 does, even though Section 230 should have been limited to the hosting and broadcast of the content (not the recommendation/endorsement of it).
Yes, I understand. I understand Section 230 and the First Amendment.
Imagine two scenarios:
1. The Meta corporate entity posts a call to arms to kill the president. It is algorithmically shown at the top of every user's feed.
2. Famous actor Tom Cruise posts a call to arms to kill the president. It is algorithmically shown at the top of every user's feed.
In the first scenario, Facebook would be liable for inciting violence. In the second, it would not be. The difference is that Facebook is not responsible for anything posted by Tom Cruise.
That effectively allows Facebook to "speak" with impunity as long as they use Facebook users as volunteer sockpuppets.
This is only possible because of Section 230. Otherwise they'd be liable for any illegal speech on the platform, regardless of whether the algorithm spreads it or not.
