The Center for Digital Technology (CDT) has released a report titled "Moderating Quechua Content on Social Media," examining the challenges faced by social media platforms in moderating content in non-English languages, particularly indigenous and low-resource languages like Quechua. The report highlights the lack of research and resources available for effective moderation systems in these languages.
"While there is important research on content moderation systems (or trust and safety systems) as it applies to Western contexts and languages, there is much less research on languages in the Majority World, particularly those that are 'low-resource'...and indigenous languages," states the report.
Quechua, with approximately 9-10 million speakers across South America, serves as a focal point for this study. It represents one of the most widely spoken indigenous language families in the region but remains under-studied regarding content moderation. The research includes case studies from other low-resource languages such as Mahgrebi Arabic, Kiswahili, and Tamil.
The CDT employed a mixed-method approach involving online surveys of frequent social media users from Ecuador, Bolivia, and Peru between December 2024 and March 2025. Interviews were conducted with Quechua content creators, language advocates, former content moderators, and social media platform representatives. Additionally, round-table discussions included natural language processing researchers focusing on Quechua.
Key findings indicate that Quechua social media users face numerous issues when posting online compared to Spanish-speaking users. These include unjust content removals and algorithmic discrimination attributed to technological limitations of automated tools like large language models (LLMs). Experts suggest that LLMs are not yet equipped for effective moderation in Quechua due to insufficient training data.
"Quechua social media users most frequently use Facebook...with a combination of Quechua and Spanish," notes the report. Despite Facebook's popularity among this group due to its networking benefits within their community, challenges persist.
Moreover, harassment targeted at individuals posting in Quechua is prevalent; gender-based violence against women identifying as Quechua online is also noted as an issue by interviewees.
The CDT report underscores significant gaps in current automated moderation technologies and processes tailored for Quechua users. It emphasizes that existing policies are often adaptations from English or Spanish frameworks without adequate consideration for linguistic nuances unique to Quechua.
###