Will Elon Musk’s lax Twitter moderation help ignite global violence?

ADVERTISEMENT

Will Elon Musk’s lax Twitter content moderation help ignite violence around the world?

Elon Musk’s decision to gut Twitter’s content moderation capacity is poised to hit hardest in countries where English is not the main language — opening the door to human rights abuses, along with hate speech and misinformation.

Since the billionaire took over Twitter on Oct. 28, he has eliminated most of its Trust and Safety team, ceased enforcing its covid-19 misinformation policy and reneged on promises to create a content moderation council. He’s also reinstated banned accounts, including that of former president Donald Trump — in that case, based on a single Twitter poll.

Civil society and human rights groups in countries like India, Poland and Nigeria say they’re frustrated that media attention and public debate over Musk’s dramatic reshaping of Twitter has largely focused on its effects in the United States, such as the restoration of Trump’s account. In a matter of weeks, Musk has reversed hard-fought gains in the site’s ability to moderate content in languages other than English in places where the social media site has previously been used to incite violence against minority groups or subvert protest movements.

“Since Musk has taken over Twitter, this situation has become perilous,” said Wendy Via, president of the Global Project Against Hate and Extremism. “I’m sure you’ve seen the reports of how the U.S. has seen a sharp increase in hate speech, and that’s in English. Imagine what the other countries are experiencing. And banned users are now evading the system by setting up new accounts often successfully because there are no global staff to report violations and take action.”

ADVERTISEMENT

In India, Twitter had been working with civil society groups since 2018 to address “dangerous genocidal hate speech on the platform,” said Thenmozhi Soundararajan, executive director of Equality Labs, a Dalit civil rights organization dedicated to ending caste apartheid, gender-based violence, Islamophobia, white supremacy and religious intolerance.

Despite those efforts, Twitter was used to help incite deadly sectarian riots in New Delhi in 2020, in which at least 20 people died after Hindu mobs targeted Muslims. Soundararajan is now worried that any progress made in moderating tweets in India’s multitude of native languages is lost — opening the door to future misuse and even violence.

“We have an American company who is operating in a genocidal market without any oversight,” she said. “That is enabling dangerous speech in an incredibly accelerated way, because they have now basically taken out all of the moderation that was preventing even more copious hate speech.”

That has left civil society groups that had been working with Twitter and its Trust and Safety Council of external experts “sending emails to emails that no longer exist,” she said.

Countries bring unique moderation problems

Musk — a free-speech absolutist — has not publicly communicated a change in moderation standards, outside of reinstating previously banned accounts and declaring that “negative/hate” speech will be “deboosted.” He has not addressed firings and layoffs within the Trust and Safety team nor his apparent failure to make good on his promise of a content-moderation council.

ADVERTISEMENT

Twitter, whose staff numbered an estimated 7,500 people when Musk took over, is now thought to be as small as 2,500 people or less. Groups outside the U.S. that used to flag dangerous or hateful content to Twitter — particularly if that content had a unique cultural context that might be missed by the site’s U.S. employees — now don’t know who to contact.

Rosemary Ajayi, lead researcher at Digital Africa Research Lab, said that she was frustrated by how much of the conversation about Musk’s Twitter have been focused on the United States.

Nigerian Twitter users’ experiences already differed widely from that of the average U.S. user. Some people in Nigeria paid thousands of dollars for verification and Twitter allowed accounts that were engaging in widespread coordinated and inauthentic behavior to stay up for years, according to Ajayi. The latest version of Musk’s new $8 a month verification system isn’t going to stop that, she said.

While it’s too early to say there has been an increase in abusive tweets and tweets with human rights implications in Nigeria’s Twitter-sphere since Musk took over, “we have seen that people are emboldened and people aren’t very public about the fact that now that Musk is here, we are free to speak — nobody’s going to get rid of us now,” Ajayi added.

Via said the Global Project Against Hate and Extremism had seen a global increase in banned users evading those punishments, getting back on the platform without being caught.

Meanwhile, in Poland

When Russia invaded Ukraine this spring, Ukrainian refugees fled to Poland — and Polish far-right groups in the country mobilized to target them, specifically nonwhite refugees, in the border area. In some cases this led to physical attacks.

Twitter was one tool that far-right groups and politicians used to mobilize against refugees, said Rafal Pankowski, an associate professor at the Institute of Sociology of Collegium Civitas in Warsaw and head of “Never Again” Association’s East Europe Monitoring Center, which combats racism and discrimination.

Pankowski said the “Never Again” group reported various leads they saw on Twitter to the company in the early days of the war, but at that time, “Twitter didn’t understand the situation.” In May, the group met with Twitter in Dublin, at a meeting organized by the European Commission, in which Twitter representatives said they wanted to keep in touch to help stop any future attempts to mobilize hate or violence using the site.

But Twitter has not followed through on that promise, Pankowski said, even in the weeks before Musk took over. Poland’s far right has begun organizing another mobilization under the slogan, “Stop the Ukrainization of Poland,” and with an accompanying hashtag, #StopUkrainizacjiPolski.

On Oct. 15, the “Never Again” Association messaged Twitter noting that it had reported several tweets with the hashtag that the site let stand.


ADVERTISEMENT

“The hashtag [‘Stop ukrainization of Poland’] is a part of a far-right campaign against the presence of Ukrainian refugees in Poland with a clear aim to incite hostility and violence,” the group wrote. Twitter later met with the group and said it would not take action.

In subsequent conversations, the most recent at the International Network Against Cyber Hate meeting on Nov. 3, Pankowski said the “Never Again” group received the same response from Twitter — “we don’t consider the hashtag a reason to react.”

“The response overall was has been quite disappointing,” said Pankowski. “This is precisely the interfacing the online [world] and the street-level reality that we are really interested in as an anti-racist organization.”

Twitter did not respond to a request for comment. Meanwhile, nationalist groups carried banners featuring the slogan at the annual Polish Independence Day march on Nov. 11. And the EU, which counts Poland as a member, has since warned Musk that Twitter could face a potential ban in the bloc over content moderation concerns.

A top Ukrainian official also expressed concern over Twitter’s direction this month at a meeting in Washington, D.C.

ADVERTISEMENT

“We are worried that Twitter will become a major source of manipulation,” Ukraine’s Deputy Prime Minister Olha Stefanishyna told reporters at the Halifax International Security Forum. “We see that Elon Musk is using it just to test the manipulation limits. And we are really afraid that this could become a major trend so that it cannot be treated as a source of information.”

Joshua Keating contributed reporting. Thanks to Alicia Benjamin for copy editing this article.

  • Benjamin Powers
    Benjamin Powers

    Technology Reporter

    Benjamin Powers is a technology reporter for Grid where he explores the interconnection of technology and privacy within major stories.