Antisemitism on Twitter has more than doubled since Musk takeover
Number of antisemitic tweets doubles between the period immediately leading up to Musk takeover, and that since.
Research published earlier this week by the Institute for Strategic Dialogue shows that antisemitism shot up when Musk took over in October. Far from proving a temporary phenomenon as was claimed at the time by the former head of Safety and Integrity, the increased volume of antisemitic tweets has proven sustained.
The key finding of the research is that in the run up to the Musk acquisition between June and October 2022, Twitter recorded a weekly average of 6,204 tweets that were deemed "plausibly antisemitic." This means that at least one reasonable interpretation of the tweet falls within the definition of antisemitism provided by the International Holocaust Remembrance Alliance. Between October and early February the weekly average of plausibly antisemitic tweets had reached 12,762, a rise of 105%.
The researchers used AI to compile a list of words and phrases that were likely to be used as part antisemitic hate speech, and then parsed every tweet through the model. Whilst they acknowledge that no such model is perfect, they estimate that the correct decision was made 75% of the time.
After this initial exercise the researchers ran another algorithm to try to identify the main themes of antisemitism. These themes were often centered on conspiracy theory, Israel or white supremacy. Investor and philanthropist George Soros got his own category. There was also a category focused on Ukraine, and it's president Volodymyr Zelensky.
Content moderation now heavily automated
Elon Musk is, by his own definition, a "free speech absolutist." Shortly after taking over at Twitter, Musk disbanded the Twitter Trust and Safety Council, which was a voluntary body set up in 2016 to advise the site on content. He also reinstated accounts which had previously been banned, although these accounts in themselves do not account for the increase in such hateful content. Plenty of the identified tweets came from new accounts which were created round about the time Musk took over.
Twitter has also laid off thousands of employees, including many responsible for content moderation. More staff have quit the company of their own accord. Content moderation is now heavily automated, so much so that the EU warned Musk earlier this month - for a second time - that it has doubts about Twitter's ability to comply with the Digital services Act which will be enforced from next year. Hate speech is one of the categories of content that tech companies will be required to monitor their platforms for.
This spike in online antisemitic hate speech comes amid an accompanying rise in offline threats to Jewish communities, and that includes the UK and Europe as much as the US
Computing says:
Understanding the role that social media plays in real life behaviour isn't an exact science. It's difficult to truly understand the extent to which online behaviour determines what a keyboard warrior will do outside the confines of a screen.
Nonetheless, the proliferation of content, often from some very high profile accounts, which push shadowy conspiracies and promote suspicion at best and encourage outright violence at worst, really can't be viewed as anything other than terribly harmful.