Twitter Is Not Legally Responsible For The Rise of ISIS, Rules California District Court (theverge.com)

An anonymous reader quotes a report from The Verge: A lawsuit accusing Twitter of providing material support to ISIS has been dismissed by a California District Court. First filed in January, the lawsuit argued ISIS's persistent presence on Twitter constituted material support for the terror group, and sought to hold Twitter responsible for an ISIS-linked attack on that basis. Filed by the family of an American contractor named Lloyd Fields, the lawsuit sought damages from an ISIS-linked attack in Jordan that claimed Fields' life. The plaintiff's initial complaint alleged widespread fundraising and recruitment through the platform, attributing 30,000 foreign actors recruited through ISIS Twitter accounts in 2015 alone. The judge assigned to the case was ultimately not swayed by that reasoning, finding that the plaintiffs had not offered a convincing argument for holding Twitter liable. The plaintiff will have the chance to submit a modified version of the complaint within 20 days of the order, the second such modification ordered by the judge. The report adds: "Apart from the private nature of Direct Messaging, plaintiffs identify no other way in which their Direct Messaging theory seeks to treat Twitter as anything other than a publisher of information provided by another information content provider," the ruling reads. At the same time, even the private nature of Twitter's Direct Messaging feature "does not remove the transmission of such messages from the scope of publishing activity under section 230(c)(1)."

Comments

Re:Two possibilities

[SvnLyrBrto]

Or:

3. Twitter is such a massive service, with a continually growing user base, with so many tweets posted and so much data ingested; that to stop ISIS using it is an extraordinarily difficult problem that they haven't solved.

Imagine they tried to do it automatically. How do you program that AI? It has to distinguish between actual ISIS posts and accounts and actual people wishing to join ISIS versus: government types investigating ISIS, other terrorist groups claiming to be ISIS, people trolling by pretending to be ISIS, people trolling ISIS by pretending to want to join, people reporting on ISIS, people mocking ISIS, people discussing ISIS or news about ISIS, people into egyptology talking about the goddess Isis, Archer fans talking about the fictional ISIS, actual people named Isis, and so on, all without generating false positives and removing the posts or accounts of anyone besides the actual terrorist ISIS.

Block by IP address blocks that they believe are owned by ISIS? Laughable. I shouldn't need to say more.

Do it manually? Right off the bat, you need multilingual staff fluent in at least English and Arabic, plus probably French, Spanish, and Farsi. And not just academically fluent as in "I took classes in school"; but actually fluent and nuanced enough to understand the aforementioned differences and never have false positives. Those people don't come cheap, even if the task is simple and easy. They especially don't come cheap in the Bay Area, where they can probably get a better job doing software localization. But assuming you can find the talent pool of fluent multilingual willing to do the crap job of slogging through twitter looking for ISIS posts and accounts; twitter ingests half a billion tweets per DAY. I can't begin to imagine how many people it would take to screen all that; nor just how awful that job would be.