The number of English-language posts containing hate speech on X (formerly Twitter) increased by around 50% in the months following Elon Musk’s acquisition of the platform, according to a study by researchers from California. The average number of likes on such posts also rose by 70%. While the researchers acknowledge they cannot prove a direct causal link, they emphasize that the data contrasts sharply with Musk’s public promises to reduce hate speech and bot activity on the platform.
“Hate speech includes expressions of contempt toward individuals based on identity traits, such as gender, race, or sexual orientation,”
explained Daniel Hickey, PhD candidate at the University of California, Berkeley, in an interview with Newseria.
“Inauthentic activity refers to bot-like accounts—those that post suspiciously repetitive content or appear to be part of coordinated campaigns, possibly orchestrated to sway public opinion.”
Musk’s Takeover and the Rise in Hate Speech
Musk completed the acquisition of Twitter in October 2022. To assess the impact, the researchers analyzed English-language posts from January 2022 to June 2023—about 31% of all posts. They aimed to verify whether Musk’s pledges to combat hate speech and eliminate bots translated into measurable outcomes.
The results, published in PLOS One, revealed a significant surge in hate speech beginning just before the acquisition and continuing through May 2023. On average, hate speech posts increased by about 50% weekly.
“While the number of general English-language posts also rose, it did so by only 8%,”
Hickey noted.
“The surge in hate speech posts clearly outpaced the platform’s overall growth. We also found a 70% increase in likes on hate-filled posts, with transphobic content seeing the most dramatic rise—up by 260%.”
No Clear Link to Policy Changes
The researchers caution that they could not directly attribute these trends to specific policy changes made by X under Musk’s leadership, largely due to limited transparency around content moderation.
“We don’t have evidence that the platform’s algorithms reduced the visibility of hate speech posts,”
Hickey said.
“It’s difficult to study algorithmic impact, and we didn’t attempt to in this paper. But we can confirm that hate speech posts received more likes and remained visible. We need more transparency from the platform to truly understand what changes were made and how they influence content.”
Bot Activity Remains Unchecked
In addition to hate speech, the study examined bot-like activity on X. The researchers found no clear evidence of a decrease in bot accounts following Musk’s takeover.
“In fact, if anything, the data suggests a potential increase,”
Hickey said.
“Musk had promised to eliminate bots and spam accounts, but we haven’t seen any indication that those efforts have had a significant effect.”
Hickey emphasized that inauthentic accounts pose serious risks, including fraud, disinformation campaigns, and electoral manipulation. Over time, they can deepen societal divisions.
Societal Consequences
“Hate speech pits groups against each other. Coordinated disinformation further polarizes public opinion,”
Hickey warned.
“To tackle global challenges like climate change or poverty, we need cooperation, not division. Continued societal polarization undermines our ability to work together and find solutions. That’s why it’s crucial to monitor the behaviors that fuel division and develop tools to counter them.”
In conclusion, while the direct role of Musk’s policies remains uncertain, the platform’s early months under his leadership have seen measurable increases in hate speech visibility and engagement, casting doubt on the effectiveness of promised reforms.
