- Bloomberg analyzed nearly 400 posts that shared disinformation about the Israel-Hamas war.
- It reported Community Notes typically took 7 hours, and sometimes as long as 70 hours, to flag the posts.
- Repeat offenders subscribed to X Premium exacerbate the problem.
X’s Community Notes feature typically takes seven hours to flag disinformation, according to a Bloomberg analysis of posts about Israel and Gaza.
Since taking over the company formerly known as Twitter, Elon Musk has championed Community Notes as a way to give readers important context without having to remove posts.
This program relies on thousands of volunteers to flag posts that lack context or are wrong.
Users who volunteer to become contributors can suggest and rate notes to add context to a post. A Community Note only appears next to a post if contributors with a “diversity of perspectives” agree that a post needs one.
Bloomberg analyzed nearly 400 posts that contained false information from between October 7 — when news of Hamas’ attack on Israel first broke — and October 21. These were drawn from a public database of Community Notes.
“We consider posts to contain “false information” if they include provably inaccurate text, photos, or other media,” Bloomberg said.
It found that it took almost three days for some of these posts to be flagged by the Community Notes program.
Examples of the disinformation included claims that video game footage was of the Israel-Hamas war and an image of chemical attack victims in Syria from 2013 which was falsely labeled as Palestinian babies killed by Israel.
According to Bloomberg, there were also several repeat offenders of accounts that spread disinformation that were previously suspended by Twitter but reinstated when Musk took over the company. Some of these accounts’ posts were also more visible because they are subscribed to X Premium.
Musk announced last month that creator posts that have been corrected or amended by Community Notes are not eligible to make money from X’s creator compensation program. But he said they would generally stay on the platform.
He undid much of the platform’s previous system which prevented the spread of potentially problematic tweets before they could be checked by its trust and safety team — which was largely disbanded soon after Musk’s takeover.
X says that images that have been flagged by Community Notes will also receive the added context when they are duplicated and reposted, but Bloomberg found this didn’t take place for the image from Syria.
X did not immediately respond to Business Insider’s request for comment.
Read the full article here