The HRW Report: Facebook and TikTok Accused of Enabling Child Recruitment

Human Rights Watch has accused Facebook and TikTok of failing to detect and remove content that armed groups in Colombia use to recruit children into their ranks. In a report released on Monday, the organization says the platforms' algorithms "seem even to promote" such material, allowing posts that glorify armed life and advertise false job opportunities to reach young users.

According to HRW, at least 1,500 minors have been recruited since 2021. The armed groups use children to fight, gather intelligence and operate drones. Nearly half of the affected children are indigenous, and girls represent 40% of cases, often facing sexual violence. The report documents posts showing cash and jewellery, as well as videos from camps set to narcocorridos, Mexican-style songs that celebrate organized crime.

Users were also found asking in comments how to join groups such as the Clan del Golfo, the ELN guerrillas and FARC dissident factions that rejected the 2016 peace agreement. HRW says those dissident groups are responsible for 74% of child recruitment cases. The organization says it raised the problem with Meta, Facebook's owner, and ByteDance, which owns TikTok. Meta said it constantly seeks to improve its counter-recruitment strategy, while ByteDance said it proactively blocks illegal armed groups from using its platform; both companies said they removed content flagged by HRW.

The accusations land as Colombia's newly inaugurated president, who is less than a month in office, pushes a military offensive against armed groups with U.S. and Israeli support. The 2016 peace deal disarmed most of the FARC, but dissident factions and drug-trafficking groups have expanded since then.

Why Meta and ByteDance's Moderation Gaps Matter in Colombia

What the evidence shows about algorithmic amplification

HRW's claim goes beyond hosting: it says the platforms' recommendation systems appear to actively surface recruitment content. That matters because a reactive removal policy would not address the core problem. If a video showing cash, jewellery or narcocorridos is boosted to users interested in those topics, it can find new recruits before moderators act. The report's documentation is strongest on discoverability, but HRW does not publish raw algorithmic data, so the exact amplification mechanism remains an inference rather than a proven result.

Where Meta and ByteDance's responses fall short

The two companies' statements focus on removal after content is identified and on their general policies against illegal armed groups. They do not directly address HRW's central charge that algorithms promote the material. Meta's pledge to "constantly improve" and ByteDance's claim of proactive blocking are not accompanied in the report by published takedown metrics or evidence of demotion. This gap leaves open the question of whether the platforms are treating the crisis as a content-enforcement problem rather than a safety-by-design problem.

Why the Colombian conflict raises the stakes

The report lands in a context where the state is escalating military action against the same groups HRW identifies as recruiters. The fact that 74% of cases are attributed to FARC dissidents, and that indigenous children and girls are disproportionately affected, means the platform issue is inseparable from a protection crisis in rural and conflict-affected areas. For Meta and ByteDance, the report creates a public record that could be used by Colombian authorities, international bodies or advertisers demanding more rigorous safeguards.

Practical Next Moves for Platforms, Advertisers and Child-Protection Agencies

  • Trust and safety teams at Meta and ByteDance should prioritise Spanish-language Colombian recruitment content — especially posts showing cash, jewellery, camps and narcocorridos — because those are the formats HRW says evade detection and reach minors.
  • Investors and advertisers should ask Meta and ByteDance to publish Colombia-specific takedown and demotion data, rather than accepting general assurances, since the report does not show evidence of proactive algorithmic demotion.
  • Advertisers running brand campaigns in Latin America should review adjacency controls for content that glorifies armed life, because HRW's documentation suggests such material remains discoverable.
  • Colombian child-protection agencies should target interventions in regions with heavy FARC dissident presence, given the report's finding that dissident groups account for 74% of recruitment cases; outreach should also be tailored to indigenous communities and girls, who are overrepresented among victims.

Risk & Opportunity Assessment

Commercial RiskMediumThe report could intensify advertiser brand-safety concerns in Latin America if armed recruitment content remains discoverable next to mainstream ads, though no direct revenue impact is documented.
Competitive RiskLowThe allegations do not describe users shifting between platforms or a competitive advantage, but any future regulatory divergence could affect both companies' content operations.
Regulatory RiskMediumThe HRW findings create a public record that Colombian authorities or international regulators could use to demand stronger content controls; no specific enforcement action is identified in the article.
Reputation RiskHighBeing linked to content that recruits minors — including indigenous children and girls facing sexual violence — is a severe reputational risk, amplified by HRW's claim that algorithms promote the material.
Technology DisruptionLowThe issue concerns existing recommendation and moderation systems, not a new technology displacing Meta or ByteDance.
Commercial OpportunityLowThe only clear opportunity is improved safety infrastructure and transparency, which could strengthen trust but is not quantified in the report.