On the frontline of international digital labour, TikTok moderators are starting to join hands across the world as they seek protection from a brutal workplace environment where trauma and violence are par for the course.

Illustration by Lynda Ouazar

In a bright office in Jakarta, a TikTok content moderator watches a livestream shift from ordinary to traumatic within seconds. A car crashes into another. The sound overwhelms the feed. Blood covers the ground while a woman kneels beside someone who no longer moves.

For moderators working on live content, decisions must be made in real-time while the stream continues. Official company guidance tells workers to step away and take time to recover if harmful content affects their well-being. In practice, stepping away lowers the number of videos moderated that day. Lower output affects performance scores, which influence pay, bonuses, and job security. Moderators describe this as an unspoken trade-off between protecting their mental health and meeting targets.

Across TikTok’s global moderation hubs, this tension shapes the job. Interviews with current and former moderators across multiple regions describe a workforce that is expected to absorb trauma at speed while remaining largely invisible to users and absent from the company’s public story.

In Johannesburg, a moderator sits in an open-plan office reviewing grainy footage of teenagers self-harming in a school bathroom. The clip is not live, so the moderator watches the video a few times to find violations before tagging policies accordingly. A few metres away, a colleague assigned to the same regional queue scrolls through cooking videos and dance trends. She finishes her shift tired, but unaffected. Both moderators face the same productivity and quality targets. Moderators in different countries describe this contrast as a normalised imbalance. Two moderators may sit side by side while carrying different psychological loads. Performance metrics make no distinction; productivity remains the primary measure of value.

The physical layout of TikTok’s moderation offices reinforces this. Offices in Casablanca, Dublin, Warsaw, Kuala Lumpur, Bucharest, Istanbul, Hanoi, and other locations follow the same blueprint. Rows of desks sit inches apart. Screens face outward. Headsets remain on for long hours. Privacy is minimal. Exposure extends beyond a single worker’s queue.

Moderators also describe a structural contradiction within TikTok’s Trust & Safety operations. Workers say employees from other departments are generally not permitted to enter Trust & Safety floors without senior approval, reflecting an acknowledgement that the content reviewed there carries a risk of harm. These restrictions exist to protect staff who are not expected to encounter graphic material.

At the same time, moderators assigned to highly traumatic queues often work in the same physical spaces as colleagues reviewing lighter content within the Trust & Safety department itself. Workers say those handling livestreams or violent material sit alongside colleagues in softer queues, with little separation and identical pay structures, despite vastly different exposure levels. The result, they argue, is a system that formally shields non-moderation staff from harm while normalising unequal risk among moderators themselves.

Moderators assigned to less severe content say they can still see or hear disturbing material from nearby screens. Audio leaks through shared spaces. A glance up reveals graphic images not meant for their queue. Workers describe an environment where emotional containment becomes difficult, even for those assigned to cleaner queues.

Administrators in livestream queues report the most intense exposure. They encounter violent accidents, sexual exploitation, self-harm, pornography, child endangerment, suicide attempts, and crimes unfolding in real time without warning. They cannot slow the stream or intervene. They classify according to the latest policy updates and move on. Several describe headaches, ear pain, blurred vision, neck strain, sleep disruption, and depression layered on top of emotional exhaustion.

Enforcement still feels inconsistent to many workers. Staff say disturbing content often remains online when internal systems fail to act, leaving logged violations unresolved for extended periods. Content removed in one region may remain untouched in another, and workers recall being held responsible for decisions the system later contradicts.

The International Language of TikTok

This all reflects the international structure of TikTok’s moderation system. The company presents itself as a unified global platform, yet moderation runs through segmented regional queues shaped by shifting policies. A gesture treated as sexualised content in Europe may be allowed in Latin America. Twerking, for example, is often flagged in European markets while remaining permissible elsewhere. Smoking triggers a ban on some days and only a warning on others, then returns to a ban weeks later, reflecting frequent changes in internal guidance driven by business priorities and algorithmic considerations, more than by health concerns. Moderators who switch between queues describe regulatory whiplash, where the same behaviour moves between violation and compliance depending on market priorities.

Policy changes arrive frequently. Some employees report updates weekly, or even daily. Others describe instructions changing within a single shift. Workers link these shifts to engagement goals, advertising sensitivities, and public scrutiny during major events, not only user safety.

This instability undermines confidence in the system. Moderators say the job often feels tied to platform optics. Some report pressure to treat borderline content leniently when engagement is high, including inconsistent age assessments of teenage content depending on the queue and market focus. These accounts suggest moderation serves business priorities as well as safety goals.

Beyond exposure, moderators describe feeling undervalued inside the workplace. Across regions, many describe the role as low status and easily replaceable, despite its essentiality to the platform’s operation. Some report dismissive treatment from colleagues in policy, product, or engineering teams, who treat moderation as unskilled or temporary work.

This perception sits uneasily with recruitment standards. Job adverts often list a requirement for a bachelor’s-level education and language fluency. In practice, many moderators enter the role from retail, hospitality, or call centre work. Some hold degrees. Many do not. Workers recount a gap between the stated requirements and how the role gets staffed, with responsibility high and status low.

Pay reinforces the stigma. Moderators often sit in the lowest pay bands in the workforce at TikTok. Many describe feeling marked by their wages, aware that salary signals position at the bottom of the hierarchy. Workers describe a social divide inside offices, where those handling harmful content feel looked down on by colleagues who benefit from the safety their work provides.

Salary gaps across this workforce are stark. Moderators based in London report salaries around £24,000 a year. In Turkey, however, some people report wages of around £525 a month. In Morocco, this falls to an estimated £450 a month. In parts of Africa, moderators report rates between £1.40 and £2.50 per hour, depending on language and contract. These shockingly low wages prove that moderator employees are typically severely underpaid for their traumatic labour, though their work is deemed crucial.

Two-Tier Labour

Moderators occupy the lowest tier of TikTok’s labour hierarchy. Their work keeps the platform usable for advertisers, creators, and users. Yet they remain among the least protected and lowest paid. Digital labour researchers argue that this inversion exposes a core feature of the platform economy: profit depends on workers who absorb harm before it reaches the public.

Moderators also describe limited access to development and progression compared with other TikTok employees. While staff in engineering, policy, or commercial roles often move between teams, attend international offsite meetings and trainings, or apply for internal transfers, moderators say similar opportunities remain rare. Several describe moderation as a terminal role, with few clear pathways into other parts of the company.

This lack of mobility affects motivation and retention. Moderators report watching colleagues in higher status roles gain visibility, training, and international exposure, while their own work remains confined to the same queues and tighter targets. Workers say internal transfer processes exist in theory but remain very difficult to access in practice. The result, they argue, is a two-tier workforce in which those absorbing the most harm face the narrowest prospects for advancement.

The global distribution of labour deepens this divide. TikTok maintains prominent offices in cities such as London, Los Angeles, Paris, Amsterdam, Tokyo, Singapore, and Dubai. These sites house engineering, advertising, policy, and strategy teams. Moderators report that large-scale content moderation increasingly concentrates in lower-paid regions, where labour costs less.

Time pressure intensifies the damage. Productivity is measured in videos per minute. Slowing down after distressing content reduces accuracy scores. Falling below quota triggers warnings, performance plans, or termination. Moderators describe emotional labour being treated like factory output, with no formal allowance for shock or recovery.

Targets also rise over time. Handling speeds shorten while volume expectations increase. Some moderators report requirements to classify more than a hundred short videos per hour, even when the content is complex. As platforms expand, workloads rise without matching support. Workers describe attrition framed as efficiency.

The lack of collective representation deepens vulnerability. Moderators in multiple regions report limited union recognition and weak routes to independent representation. Workers say this leaves them isolated when raising concerns about workload, mental health, pay, or unfair treatment. Labour advocates argue that fragmentation across countries and employment statuses limits collective power, making it harder for workers to assert rights.

The psychological toll appears repeatedly in worker accounts. Moderators describe anxiety, depression, dissociation, intrusive imagery, and emotional numbness. Counselling services, when available, are often limited. Several workers say that seeking support affects performance assessments, discouraging disclosure. Advocacy groups describe this as a systemic failure to protect workers whose jobs require sustained exposure to trauma.

Radical Moderation

Technology executives often argue that AI will replace human moderation. Workers and researchers dispute this. Automated systems struggle with context, sarcasm, cultural nuance, and age assessment. They often fail to distinguish staged harm from real danger. Moderators argue that automation helps to obscure human cost while keeping wages low and protections weak.

The consequences extend beyond the workforce. When moderation fails, misinformation spreads, extremist content circulates, harmful challenges escalate, and minors face greater risk. Researchers describe moderation as an informal layer of public safety infrastructure, shaping what millions see and what gets suppressed. Yet the people doing this work often operate without the protections that frontline roles usually receive.

Resistance is growing. Moderators in Kenya have taken technology companies to court. Workers in Ireland, Germany, and the UK have sought union representation. Digital rights organisations increasingly classify moderation as trauma labour and call for stronger regulation. A transnational workforce is beginning to organise across borders, linked by shared experience rather than shared geography.

Moderators protect children. They intercept violence. They limit exploitation and slow the spread of hate. They absorb harm that would otherwise reach millions. Yet they remain among the least protected workers in the digital economy. TikTok does not function without them, and neither does any major social platform.

A safer digital public sphere requires recognising this invisible frontline. Their labour deserves dignity, protection, and fair compensation. Without structural change, platforms will continue to rely on workers who carry the heaviest burdens while receiving the fewest safeguards. Their work needs visibility, value, and defence, not silence from systems that depend on it.