Kirk here. We've posted about this topic before, but this article provides new information (at least new to me) about the nature of the role of X and X's algorithms / AI in the spread of hateful nonsense.
X seems to be significantly worse than Facebook in regard to the spread of misinformation like this but the Facebook algorithm has its problems too. One issue is that images with text tend to spread better than images and text with a citation about the source of the information. One illustration of how this can play out is the following post by Joe with some comparative comments about this earthquake and the 2016 earthquake.
https://www.facebook.com/1379708514190936/posts/1471486435013143
It's well written and deserves to be spread but, nonetheless, I think Facebook's analytics say as much about the Facebook algorithm as they do about the post: 433 reactions and 28 shares. Facebook is set up so that a post that primarily introduces a link can't get that kind of reaction because the linked article is immediately divided from the post at share time. I can't be sure but I suspect that if Joe had added links to relevant sources that backed his claims that the Facebook algorithm would not have promoted it as much. Here's a link to a source to back up what I am saying (though, ironically, I suspect that adding this source may reduce the "reach" of this post):
Now, in Joe's case, there's absolutely nothing wrong with lots of people seeing his post; his comments are good and based in fact. But, as a general rule, promoting information that is NOT sourced and downplaying information that is sourced strikes me as a recipe for the spread of misinformation.
No wonder we see MORE misinformation in 2026 than in 2016.
--- start quote ---
When The Yomiuri Shimbun examined X posts and reposts containing terms such as “earthquake” and “pork” using social media analysis tool Social Insight, the number of posts surged to over 20,000 on July 29, the day after the earthquake. Despite subsequently declining, the number exceeded 10,000 on some days in August.
X has a feature that uses artificial intelligence to summarize attention-grabbing posts and shows the summaries to users. In early August, it even displayed misinformation that Muslims were complaining about hot meals provided following the Kumamoto Earthquake.
--- end quote ---
https://japannews.yomiuri.co.jp/society/general-news/20260916-348953/