A new study by the University of California in Berkeley revealed that hate speeches increased by 50 % in the months which followed the acquisition of X, formerly Twitter, by billionaire Elon Musk.
According to a new study, hate speeches increased by 50 % on the X social media platform in the months which followed its acquisition by billionaire Elon Musk.
The study of the University of Berkeley (California), published Wednesday in the journal Public Library of Scienceused artificial intelligence (AI) to randomly collect messages on X containing terms or sentences linked to hate speeches in English, since the acquisition of X (then known as Twitter) by Musk in October 2022 until June 2023, when he left his CEO post.
The researchers then sorted them to check if each message met the criteria and to get rid of duplicates or reruns.
They also collected a “control group” of messages from January 2022 to compare the amount of hatred on the platform.
The analysis revealed that hate speeches had doubled in the weeks preceding Twitter control by Musk and that they had continued until May 2023.
During this period, the number of likes on messages using homophobic, transphobic and racist insults increased by 70 %, This suggests that more users have been exposed.
“I think that the increase in hate speeches that we see is worrying because it can mean that marginalized communities do not feel welcome or in danger on the platform”, said Dan Hickey, doctoral student in Berkeley, in Euronews Next.
The report continues by indicating that the number of bots and other inauthentic accounts has not decreased during this period, despite Musk’s public promises to reduce them.
Their study follows others who noted an increase in hate speeches or anti -Semitism after the acquisition of Musk in 2022, including desrapports of the Institute for Strategic Dialogue (ISD) and the Center for the Fight against digital Lahaine.
X must be more transparent
Two peaks of hate speech occurred before Musk became CEO: one in April 2022, coinciding with a vast advertising campaign featuring a transgender woman, continues the Berkeley report.
The second occurred during the few weeks of October which preceded the control of Musk.
The report continues by indicating that a “reshuffle of management” took place shortly after the arrival of Musk at the head of the company, as well as theDissolution Advisory Council on Confidence and Security, who took care of the moderation of the content.
However, researchers say they cannot not draw “definitive conclusions” On the question of whether Musk is at the origin of the peak of hate speech on X, due to the lack of information on “Specific internal changes” of the platform.
“To be able to establish a final link, it would be necessary to know more about what has changed and when, and the organization has not been transparent on this subject”said Daniel Fessler, professor of anthropology in Berkeley and member of the research team, in Euronews Next.
After June 2023, Mr. Hickey said that X had started to invoice the data collected by his application programming interface (API) and that they no longer had the necessary budget to extend their study.
ADVERTISEMENTWhat is X’s position on hate speech?
A few days after the acquisition of Twitter, Muska declared that “The firm commitment of the company in favor of the moderation of the content remains absolutely unchanged” And that hate speeches have decreased since it owns the company.
“Freedom of expression does not mean freedom of access”, had insisted Elon Muskdans a message on X. “Negativity should and will have less access than positivity”.
X rules also indicate that users “cannot directly attack other people on the basis of race, ethnicity, national origin, caste, sexual orientation, gender, gender identity, the ‘Religious affiliation, age, disability or serious illness “ on the platform.
The lasttransparency report of the platform reveals that more than 2,000 accounts have been suspended And that a little less than 5 million hateful messages have been deleted on the 66 million complaints filed by users for hateful behavior between January and June 2024.
ADVERTISEMENTThe company also has hate speech in its category “Abuse and harassment”.
Of the 82 million complaints received, X suspended a million accounts and deleted or labeled 2.6 million messages, according to the report.
More recently, xsigned a voluntary commitment(Link in English) of the EU to combat hate speech online. This agreement requires social media companies to let experts control the way in which they examine hate speech opinions and to teach users how to report illegal hateful content on their sites.
