The Platform Question: Power, Accountability and Global South When moderation gets involved with politics In 2024, The Guardian revealed testimonies from Meta workers who denounced unfair moderation policies toward the Palestinian community.[10] According to an anonymous employee and a letter signed by more than 200 people, internal rules known as Community Engagement Expectations (CEE) are used to remove discussions about Gaza, even when they express condolences or question product errors that affect the visibility of Palestinian voices. The document contrasts the institutional silence in the face of more than 35,000 civilian deaths in Palestine with the public support the company expressed toward its Israeli employees after the attacks of October 7, 2023. According to an anonymous employee and a letter signed by more than 200 people, internal rules known as Community Engagement Expectations (CEE) are used to remove discussions about Gaza, even when they express condolences or question product errors that affect the visibility of Palestinian voices. 36 The signatories accuse a long history of biased moderation against Palestinian content, despite calls from Human Rights Watch and Senator Elizabeth Warren to ensure greater transparency. In their letter, they demand an end to internal censorship, recognition of Palestinian suffering, and a public statement calling for a ceasefire. The report also exposes the structural fragility of the moderation system: human decisions must align with Meta’s policies, but the application of those rules varies depending on language and geographic context. Moderation, therefore, is not just a technical or automated exercise, as it also faces tension between corporate standards, international treaties, and specific local contexts. And in Latin America? Returning to our experience, from 2023 to the present, we have noticed a steady increase in requests for help sent to the Amaranta NGO’s email address, not only from Chile but from various Spanish-speaking countries. Seeking to systematize the cases and trying to find the source of these requests, we discovered that an organization that had been providing services to Meta for some time recommended us as partners for cases of technology-facilitated gender-based violence. No one consulted us or informed us of this. This replicates what other civil society organizations denounce, which is how a multimillion-dollar corporation shifts responsibility to small organizations in the global south. This fact accurately sums up the extractive and unequal logic of today’s digital ecosystem. A Big Tech Accountability in Content Moderation with Feminist Perspective The Platform Question: Power, Accountability and Global South Week after week, we continue to receive cases of digital sexual violence, non-consensual dissemination of intimate images, and identity theft, including cases from Arab countries. We try to respond to each message with humanity, knowing that behind every automated form that did not work for the victims, there is a story that the platforms do not want to address. Ivana Feldfeber points out the following in the prologue to the report Helplines for addressing cases of gender-based violence in digital environments[11] : “It is essential to consider the role of technology companies, which facilitate the use of these platforms. And at this point, it is necessary to emphasize that companies do not give us free products altruistically; they are not mere service providers to society. Rather, behind these social networks are large economic interests, with millions of dollars in contracts with companies, governments, and other groups to offer advertising and traffic their users’ data (...) These companies, when asked to block users, remove content disseminated without consent, or recover hacked accounts, they often do not comply with these requests. The same companies that censor images of female nipples on their platforms ignore calls for help from their users who are victims of digital gender-based violence.” Meta, like other companies in the sector, outsources digital violence to precarious and subcontracted workers, making their suffering invisible and normalizing technological progress based on the emotional precariousness of those who moderate content and also train artificial intelligence to appear more human for the same price, without forgetting that they end up delegating their responsibility for digital violence to civil Conclusion The discussion about content moderators is becoming political and requires immediate attention. Who decides what is acceptable to see on the internet? And at what human cost is that decision sustained? Can we talk about safe digital spaces if this task is left in the hands of precarious workers with low wages and little psychological protection? How can we demand accountability from platforms that includes all these aspects if they are currently invisible due to outsourcing? Meta, like other companies in the sector, outsources digital violence to precarious and subcontracted workers, making their suffering invisible and normalizing technological progress based on the emotional precariousness of those who moderate content and also train artificial intelligence to appear more human for the same price. A Big Tech Accountability in Content Moderation with Feminist Perspective 37