A new study from the University of California, Berkeley, reports that hate speech on X (formerly Twitter) increased by 50% following Elon Musk’s acquisition. Researchers used AI-driven analysis to review posts between October 2022 and June 2023, revealing spikes in homophobic, transphobic, and racist content. Despite Musk’s promises to reduce bots and harmful content, the findings suggest that hate speech increased and gained higher engagement rates.
The study aligns with prior research from the Institute for Strategic Dialogue (ISD) and the Centre for Countering Digital Hate, documenting a rise in toxic online discourse post-acquisition. While the researchers admit that internal platform changes remain opaque, their analysis points to a moderation and content visibility shift.
What’s Happening & Why This Matters
AI-Powered Analysis Reveals Hate Speech Surge
Researchers at UC Berkeley leveraged AI algorithms to assess X posts containing hate speech-related keywords:
- The study analyzed data from Musk’s takeover in October 2022 through June 2023, tracking harmful content engagement and visibility fluctuations.
- Pre-takeover data from January 2022 provided a comparative baseline, helping to measure the increase in explicit hate speech.
- Hate speech doubled leading up to Musk’s acquisition, reaching its highest levels in May 2023.
The study also found that hate-fueled posts attracted higher engagement:
- Likes on posts containing slurs increased by 70%, reinforcing that harmful content gained algorithmic priority.
- Content visibility grew despite Musk’s claims of de-ranking harmful speech, raising concerns about the platform’s moderation effectiveness.
- Despite efforts to crack down on bot accounts, researchers found no measurable reduction in inauthentic engagement.
X’s Content Moderation Shift Under Musk
Following Musk’s acquisition, X underwent rapid leadership and structural changes:
- The Trust and Safety Advisory Council, responsible for platform moderation, was disbanded entirely.
- Policy transparency declined, making monitoring changes increasingly difficult for external watchdogs.
- Musk publicly stated that X’s content policies had not changed, though external analyses suggest a significant shift in enforcement and algorithmic promotion.
While researchers refrain from directly attributing the rise in hate speech solely to Musk’s leadership, they suggest that weakened moderation likelamplifiedng harmful discourse.
X’s Response to Hate Speech Allegations
X continues to assert that it actively moderates harmful content:
- Over 2,000 accounts were suspended, and nearly 5 million hate speech-related posts were removed between January and June 2024.
- The company reported receiving 66 million complaints regarding hate speech, but only a small percentage resulted in removals.
- X signed a voluntary EU pledge, committing to greater oversight, enhanced moderation, and user education on digital safety.
Musk has repeatedly defended his free speech absolutist stance, claiming that harmful content is suppressed through algorithmic downranking. However, independent research suggests that harmful content continues to thrive on the platform.
TF Summary: What’s Next
The debate over free speech and content moderation on X remains heated. While Musk maintains that hate speech visibility has decreased, independent studies show a rise in its presence and engagement rates. Researchers urge greater transparency from X, arguing that open-access data is crucial for assessing platform safety. With regulatory scrutiny increasing, the future of hate speech moderation and digital governance on X remains uncertain.
— Text-to-Speech (TTS) provided by gspeech