Finally got around to re-reading the infamous Section 230, and large chunks of the Communications Decency Act to which it belongs.
Man, lots of the takes were very, very off.
Twitter is actually quite onside in its choices to flag and fact-check tweets.
Man, lots of the takes were very, very off.
Twitter is actually quite onside in its choices to flag and fact-check tweets.
Section 230 immunity covers online platforms that even if they go as far as edited posted content, as long as they do not substantially change the meaning of the content.
The platform can flag the content to its heart's content.
The platform can flag the content to its heart's content.
And thinking about the whole thing from both a Common Law and US Constitution angle, the Free Speech arguments really do not seem to apply. First Amendment protects speaker from government, not speaker from publisher OR platform.
And looking back at the debates last month, there was a false assumption that platforms have immunity from defamation for 3rd party content *in exchange* for guaranteeing free, unfettered speech by those 3rd parties. No the case.
Reading the CDA and judgments that came before it (Cubby v CompuServe, Stratton Oakmont v Prodigy), it's clear immunity from defamation is the default for all distributors, and that the narrow set of distributors called "publishers", open to defamation are seen as the exception.
Essentially, US law doesn't seem to treat you as a publisher unless you demonstrably have:
1. Realtime/advance knowledge of all content
2. Opportunity to exercise editorial control on it.
3. Ability to do so.
Clearly every social media platform admin fails 1 & 3 at least.
1. Realtime/advance knowledge of all content
2. Opportunity to exercise editorial control on it.
3. Ability to do so.
Clearly every social media platform admin fails 1 & 3 at least.
Notice the emphasis is on ALL content.
This dismisses the argument put forward by some, myself ignorantly included, that if Twitter fact-checks Trump but not others, they've automatically failed to meet the test. Partial ability to curate content in fact PROVES they qualify.
This dismisses the argument put forward by some, myself ignorantly included, that if Twitter fact-checks Trump but not others, they've automatically failed to meet the test. Partial ability to curate content in fact PROVES they qualify.
I went down this rabbithole because I wanted to tweet about Zuck's woes with Facebook's advertisers and shareholders over the platform's choosing not to police misinformation. Having re-read Section 230 and other parts of the CDA BEFORE that thread, it's content will change.
FYI if platforms eventually lose Section 230 Immunity, the most likely cause would be the rise of AI. Once AI can crawl the entire TL and flag almost every questionable post within tolerance level for human censors, the "inability" argument dies. Facebook is digging its grave. 
By "questionable" I mean specifically potentially defamatory, libellous, or slanderous. Not hate speech etc.
But of course, this implies strong fact checking ability growing alongside, which is already happening with advances in data journalism, semantic web, and NLP.
But of course, this implies strong fact checking ability growing alongside, which is already happening with advances in data journalism, semantic web, and NLP.
