The use of hateful language on Twitter has increased substantially since Elon Musk took over, according to new research.
Mr Musk completed his $44bn acquisition of the social networking platform in late October, promptly laying off roughly half of the company’s 8,000 workers.
Although the head of trust and safety, Yoel Roth, said at the time that frontline moderation staff had experienced “the least impact”, last week – having resigned – he said Twitter was not safer under Mr Musk.
On Friday, the US-based Center for Countering Digital Hate (CCDH) released a report looking at the increase in hateful language on the platform.
Among its findings:
• Daily use of the word “n****r” was up three-fold since Mr Musk’s takeover, compared to the average seen across 2022
• Daily use of “c***s” was up by 33% in the same time period
• Daily use of “f****t” was up by 58%
• Daily use of “tr***y” was up by 62%
The news comes just days after Mr Musk tweeted his congratulations to the team at Twitter, saying “hate speech impressions” were “down by one-third from pre-spike levels”.
He also claimed on 18 November that Twitter’s new policy was “freedom of speech, but not freedom of reach”, adding: “Negative/hate tweets will be max deboosted and demonetised, so no ads or other revenue to Twitter.
“You won’t find the tweet unless you specifically seek it out, which is not different from the rest of the internet.”
Kanye West suspended from Twitter after sharing image of swastika and praising Hitler
Twitter to launch gold ticks in ‘painful but necessary’ move
Elon Musk warned by European Union that Twitter could be banned unless he sticks to bloc’s digital rules
But the CCDH report found that average engagement with tweets containing slurs had also jumped.
In the two weeks before Mr Musk’s takeover, the average number of replies, retweets and likes for tweets using “n****r”, “tr***y” or “f****t” was 13.3 but after the takeover, it jumped to 49.5 – an increase of 272%.
As a case study, the research looked at anti-LGBTQ+ tweets after the Colorado Springs shooting, where five people were killed at an LGBTQ nightclub on 19 November.
They concentrated on the hateful “grooming” rhetoric, which they said was viewed on tweets tens of millions of times after the shooting.
In the three months before Mr Musk took over Twitter, the 10 accounts most responsible for that rhetoric gained 222,709 followers a month between them, taken as an average across the three months.
In the month following Musk’s takeover, they gained a total of 944,205 followers – an increase of more than 320%.
It is not the first time concerns have been raised about how Mr Musk’s leadership is changing Twitter.
The rolling back of the platform’s COVID misinformation policy is expected to give more freedom to anti-vaxxers, while his announcement of a “general amnesty” will allow a return of those previously banned for violating Twitter rules.
Among them are white supremacists, misogynists, and far-right conspiracy theorists.
On Friday, Mr Musk appeared to be addressing the CCDH report when he tweeted: “Hate speech impressions – number of times a tweet was viewed – continue to decline, despite significant user growth.”
He promised that data would be published weekly, adding: “Freedom of speech doesn’t mean freedom of reach.
“Negativity should and will get less reach than positivity.
“There are about 500 million tweets per day and billions of impressions, so hate speech impressions are less than 0.1% of what’s seen on Twitter.”