Why dangerous content is thriving on Kenyan social media

comment

NAIROBI – The gunman approaches from behind and points a pistol to his victim’s head. He pulls the trigger and “plop,” a lifeless body sags forward. The shot switches to another execution and another.

The video was posted on Facebook, among a large group of Al-Shabab and Islamic State supporters, where various versions were viewed thousands of times before being removed.

As Facebook and its competitor TikTok grow at breakneck speed in Kenya and across Africa, researchers say the tech companies are not keeping pace with the spread of terrorist content, hate speech and false information, and are exploiting poor regulatory frameworks to avoid tighter oversight.

“It’s a conscious choice to maximize labor and profit yield because they see societies in the Global South primarily as markets and not as societies,” said Nanjala Nyabola, a Kenyan technology and social researcher.

About one in five Kenyans uses Facebook, which rebranded itself to Meta last year, and TikTok has become one of the country’s most downloaded apps. The proliferation of violent and inflammatory content on the platforms poses real risks in this East African nation as it prepares for a tightly contested presidential election next month and grapples with the terrorist threat posed by a resurgent al-Shabab.

“Our approach to content moderation in Africa is no different than anywhere else in the world,” wrote Kojo Boakye, Meta’s director of public policy for Africa, the Middle East and Turkey, in an email to The Washington Post. “We prioritize security on our platforms and have taken aggressive steps to combat misinformation and harmful content.”

TikTok’s Head of Government Relations and Public Policy in Sub-Saharan Africa, Fortune Mgwili-Sibanda, also responded by email to The Post, writing: “We have thousands of people working on security around the world – and we’re still expanding this feature in our African markets in line with the continued growth of our TikTok community on the continent.”

Companies’ content moderation strategy is two-pronged: artificial intelligence (AI) algorithms provide a first line of defense. But Meta has admitted that teaching AI to recognize hate speech in multiple languages ​​and contexts is difficult, and reports show that posts in languages ​​other than English often fall through the cracks.

In June, researchers at the London-based Institute for Strategic Dialogue (ISD) released a report detailing how al-Shabab and the Islamic State use Facebook to distribute extremist content, such as the execution video.

The ISD’s two-year investigation uncovered at least 30 public al-Shabab and Islamic State propaganda sites with a combined total of nearly 40,000 followers. The groups released videos showing gruesome assassinations, suicide bombings, attacks on Kenyan forces and Islamist militant training exercises. Some content has lived on the platform for more than six years.

The dependency on AI is a core problem, said the report’s co-author Moustafa Ayad, because bad actors have learned how to play off the system.

If the terrorists know the AI ​​is looking for the word jihad, Ayad explains, they can “break up JIHAD with dots between the letters, so now it won’t read it right [the] AI system.”

Ayad said most of the accounts flagged in the report have since been removed, but similar content has since surfaced, such as a video released in July featuring Fuad Mohamed Khalaf, an al-Shabab leader wanted by the US government. It was viewed 141,000 times and shared 1,800 times before it was removed after 10 days.

Terrorist groups can also bypass human moderation, the second line of defense for social media companies, by exploiting linguistic and cultural skills gaps, the report says. Kenya’s national languages ​​are English and Swahili, but Kenyans speak dozens of other tribal languages, dialects and local slang. sheng

Meta said it has a 350-strong multidisciplinary team, including native Arabic, Somali and Swahili speakers, that monitors and handles terrorist content. Between January and March, the company claims to have removed 15 million pieces of content that violated its terrorism policies, but didn’t say how much terrorist content it thought was still on the platform.

In January 2019, al-Shabab attacked the DusitD2 complex in Nairobi, killing 22 people. A state investigation later found they planned the attack using a Facebook account that local media said went undetected for six months.

During Kenya’s last election in 2017, journalists documented how Facebook became a major source of disinformation and struggled to curb the spread of ethnically charged hate speech. They now fear TikTok has become the platform of choice for those looking to stoke tensions ahead of the August 9 presidential election.

In June, the Mozilla Foundation released a report detailing how election-related disinformation has taken hold on TikTok. The report examined more than 130 videos from 33 accounts with more than 4 million views and found ethnically based hate speech, manipulated and false content that violated TikTok’s own policies.

A video clip mimicked a laundry detergent commercial in which the narrator told viewers that the “laundry detergent” could wipe out “Madoadoa,” including members of the Kikuyu, Luhya, Luo, and Kamba tribes. Literally, “madoadoa” is a harmless word meaning blemish or stain, but it can also be a coded ethnic slur and an incitement to violence. The video contained vivid images of clashes following last year’s elections✓.

Following the report, TikTok removed the video and flagged the term “madoadoa,” but the episode showed how the nuances of language can elude human presenters. A TikTok whistleblower told the report’s author, Odanga Madung, that she was asked to watch videos in languages ​​she didn’t speak and use images alone to determine if they were violating the guidelines.

TikTok didn’t directly respond to that claim when asked by The Washington Post, but the company recently issued a statement about efforts to address problematic election-related content.

TikTok said it moderates content in more than 60 languages, including Swahili, but declined to give further details about its moderators in Kenya or the number of languages ​​it monitors. It has also set up a Kenya-specific operations center with experts who identify and remove posts that violate its policies. And on July 14th, an in-app user guide was launched with information on elections and media literacy.

“[We] have a dedicated team working to protect TikTok during Kenya’s elections,” Mgwili-Sibanda wrote. “We prohibit and remove false voting information, incitement to violence, and other violations of our policies.”

However, researchers still fear that violent rhetoric online could lead to real violence.

“You will see that these lies have really very tragic consequences for the people who take part in rallies,” said Irungu Houghton, Amnesty International Kenya director.

Researchers say that in Kenya, TikTok and Meta are getting away with lower standards for content moderation, in part because Kenyan law doesn’t hold social media companies directly responsible for harmful content on their platforms. In contrast, Germany’s “Facebook law” fines companies up to $50 million if they fail to remove “clearly illegal” content within 24 hours of a user filing a complaint.

“It’s quite a gray area,” said Mugambi Laibuta, a Kenyan lawyer. “[W]When it comes to hate speech, there is no law in Kenya that says these sites should enforce content moderation.”

If Meta and TikTok don’t police themselves, experts warn, African governments will do it for them, potentially in anti-democratic and dangerous ways.

“If the platforms don’t pull themselves together, they become convenient excuses for authoritarians to crack down on them across the continent … a convenient excuse for them to disappear,” Madung said. “And we all need these platforms to survive. We need them to thrive.”

About Sonia Martinez

Check Also

EPRA announces plans to protect Kenyans from high fuel and electricity prices

Energy Petroleum and Regulatory Authority (EPRA) Director-General Daniel Kiptoo has outlined the government’s plan to …