TikTok's Automated Moderation (is still better than no moderation)

dek: also ft; the rise and rise of Meta’s court cases

February 15, 2025

CybAfriqué is a space for news and analysis on cyber, data, and information security on the African continent.

HIGHLIGHTS

Notes on TikTok’s automated moderation

We’re currently mulling over TikTok’s Community Guidelines Enforcement Report for Q3 2024. TikTok is one of the last remaining social media platforms that still actively moderate content, the transparent report provides valuable information for stakeholders to understand the online behaviors of users and the information manipulation landscape.

Of 147 million videos removed globally, just over 12.6 million were from Africa. Egypt, Nigeria, and Algeria topped the chart in Africa with 2.3 million, 2.1 million, and 1.9 million videos respectively, and the USA topped globally with 10.5 million videos removed.

TikTok says its human-augmented automated system detected and removed 99.1% of these videos just after 24 hours. 77% of all flagged content was removed without human oversight.

Sensitive and sexually explicit videos made up the largest percentile, nearly 30%, of all total removed videos. Activists and political leaders, including Kenyan President Williams Ruto have in the past complained about underage exposure to sexual content and “erasure of African cultural values” on the platform.

In parts of Northern Nigeria, TikTokers have been arrested and charged with moral offenses.

Senegal and Somalia had earlier shut the platform due to unchecked content threatening the security and stability of the countries. TikTok content has also been linked to violent incitement and hate speech against Algerians in France, which led to the arrests of several influencers and a diplomatic rift between France and Algeria.

Meta’s rising court cases

Nigerian human rights lawyer’s Femi Falana recently filed a case against Meta. Femi Falana (SAN) files a $5m lawsuit against Meta over alleged invasion of his privacy on Facebook.

He alleged that his name, image, and voice were wrongfully used in a sponsored ad for a health product on its platform, portraying him as suffering from prostatitis, a condition Falana insists he’s never had.

In his suit, filed at the Lagos State High Court on February 3, 2025, Falana argued that the unauthorized use of his identity constitutes a violation of his fundamental right to privacy under the Nigeria Data Protection Act of 2023.

The use of celebrities or popular personalities in false adverts has been a common case on Facebook. Africa Check, a fact-checking platform, also debunked some similar ads targeted at Nigerian journalists. Pictures of these popular important personalities are being used to fake the credibility of advertised products in sponsored ads.

Meta is still fighting a fine by the Nigerian consumer protection agency for breaching Nigerians’ right to privacy and another case against ex-moderators in Kenya. Other data privacy infractions against the social media giant in Europe include the Irish Data Protection Commission fine of €251 million and a €1.2 billion fine following an EDPB binding dispute resolution decision in April 2023.

The social media giant founder, Mark Zuckerberg had earlier disclosed that Facebook’s parent, Meta, will be downsizing its global workforce to employ more AI engineers as AI takes the priority stage of the big tech.

Over 3000 employees have been receiving termination letters and severance payments since the “performance terminations” tagged layoff kicked off on Monday set to continue till 18th of February.

Per Tech Cabal, a Sub-Saharan Africa spokesperson for Meta claimed the cut was part of the company's routine performance-based layoffs that only low-performers were laid off. “We have the highest confidence in the fairness and robustness of our performance ,review process leading to these decisions” he said while noting that impacted employees are getting generous severance packages.

Do you find value in the CybAfriqué newsletter? Share to support the work we do.

FEATURES

  • About a year after he was first arrested Tigran Gambarayan has unpacked his ordeal in the hands of the Nigerian authorities. In conversation with Andy Greenberg, Gambarayan narrated what he suffered at the hands of Nigeria's Secret Police who trumped up charges against him and treated him like a dangerous felon for eight months in prison.

  • With the aid of Artificial intelligence, cases of document forgery and deepfakes are on the rise across Africa. Fraudsters are using AI to create fake documents, voices, and images that facilitate identity theft and financial crimes. This report details how cheap artificial intelligence tools are leading to a wave of biometric fraud in Africa, costing several unsuspecting victims their hard-earned money

HEADLINES

ACROSS THE WORLD

OPPORTUNITIES

Privacy Conference (PrivCon) 2025 Date: February 26, 2025

Reply

or to participate.