After months of contention between Elon Musk and Twitter executives over his bid to own the company, the Tesla founder was officially. But as the new leader emerged at the helm, the platform saw a massive spike in hate speech, a new study found.
Researchers from Montclair State University found that the 12 hours immediately following Musk’s ascension to ownership saw a much more “hostile” environment on Twitter. The team looked at tweets filled with “vulgar and hostile” rhetoric aimed at people based on their race, religion, ethnicity and orientation, such as the “n-word,” “k-word,” and “f-word,” to find out just how bad it got.
And what they found was an “immediate, visible, and measurable spike.”
In the week leading up to Musk’s acquisition, researchers said there were no more than 84 hostile tweets an hour on the platform. But from midnight on October 28 – the day Musk took ownership – to noon the next day, there were 4,778 hate-filled tweets. That accounts for more than 398 tweets an hour, about 4.7 times higher than the seven-day average leading up to that day. The potential reach for those tweets was more than 3 million, researchers found.
Researchers also said there was an increase in negative sentiment, with more than 67% of the tweets sent after Musk’s takeover having a negative tone.
“In sum, the content and tone of Twitter posts became measurably more oriented towards hate speech on the day Elon Musk became CEO of the company with significant reach attained for this hate content,” the study says.
Yoel Roth, Twitter’s head of safety and security, has also confirmed an increase in hate speech. On Monday, he tweeted that the company has seen a “surge in hateful conduct” and removed more than 1,500 accounts. A graph he shared of total impression of tweets with at least one slur shows that the increase kicked off on Friday very shortly after the Musk acquisition was announced. It appeared to spike between Saturday and Sunday.
Roth described the surge as a “short-term trolling campaign” with many of the accounts they removed stemming from repeated “bad actors.” On Saturday, he said as an example that more than 50,000 tweets that used a particular slur were issued by just 300 accounts, most of which were “inauthentic.”
“These issues aren’t new, and the people targeted by hateful conduct aren’t numbers or data points. We’re going to continue investing in policy and technology to make things better,” he tweeted.
While Montclair researchers made it clear that the acquisition saw an immediate increase in hostile language on the platform, what remains murky is what specifically caused it. Musk has long said that this ownership of Twitter would come with fewer restrictions, but researchers said it’s “speculative” to know whether his potential policy changes would have caused the spike.
It’s possible, however, that based on his past sentiments users resorted to the language because they assumed they would no longer be banned or suspended from the platform. It’s also possible that having an “unmoderated platform was potentially a source of excitement,” researchers said.
Musk has not explicitly said what will be tolerated on the platform since the surge began, although he has retweeted Roth’s statements about the surge. On Friday, Musk said that Twitter will form a “content moderation council with widely diverse viewpoints” and that no major content decisions would be made before that council can convene.
“Twitter will not allow anyone who was de-platformed for violating Twitter rules back on platform until we have a clear process for doing so, which will take at least a few more weeks,” Musk said Tuesday night. “Twitter’s content moderation council will include representatives with widely divergent views, which will certainly include the civil rights community and groups who face hate-fueled violence.”