Open Letter to
Social Media Platforms

To the leadership teams at Meta, Instagram, Facebook, TikTok, X, YouTube, and other major social media platforms,

We are writing as organisations, campaigners, advocates, journalists, survivor-support services, safeguarding professionals, and researchers working across abuse, violence against women and girls, child protection, justice, and media accountability.

Increasingly, we are seeing accounts restricted, suspended, or pushed out of view simply for using accurate language about abuse.

That should not be happening.

Across our sectors, social media has become one of the main places where we educate, campaign, support survivors, challenge harmful reporting, raise funds, and engage the public. Yet many of us now find ourselves second-guessing the language we use, because using the correct words can trigger moderation penalties.

Posts are flagged, reach drops, warnings appear, and accounts become restricted. In some cases, they disappear altogether. In most cases, that is because someone described harm plainly: sexual violence, child sexual abuse, domestic abuse, coercive control, exploitation.

At the same time, those working in this field know how easily reporting systems can be weaponised. Trolls, bad-faith actors, and coordinated harassment campaigns have learned that if they report content often enough, they are able to silence the people exposing abuse rather than those responsible for it.

That is the problem we are asking you to address.

Accurate language about abuse is not the same as abusive content

This should not need saying, but clearly it does.

There is an obvious difference between content that promotes abuse and content that explains it, reports on it, challenges it, or helps people recognise it.

When moderation systems fail to recognise that distinction, the consequences are significant. Organisations begin to soften their language. Journalists avoid certain words. Campaigners find themselves editing around the truth. Survivor advocates have to decide whether accuracy is worth the risk of losing visibility. Some continue regardless and accept the penalties. Others self-censor because they cannot afford to lose their account, their audience, or their income.

This is more than an inconvenience. It distorts public understanding of abuse and makes it harder for those working in this space to do their jobs effectively.

Language shapes understanding. It influences whether people recognise abuse when they see it, whether survivors feel believed, and whether responsibility remains with those who caused the harm instead of disappearing behind vague or sanitised wording.

When accurate language becomes harder to use, the public does not stop talking about abuse. It simply becomes harder to talk about it accurately. That serves nobody - not survivors, not journalists, not researchers, and not the platforms themselves.

If platforms make it harder to name abuse accurately, they are not remaining neutral. They are shaping the conversation around abuse, whether intentionally or not.

We are seeing the same pattern across different sectors

The organisations signing this letter do very different work. Some support survivors directly. Others work in prevention, safeguarding, journalism, law, research, or campaigning. Some focus on one form of abuse, others on several. Yet the experience is strikingly similar.

Content that uses accurate abuse-related language in a legitimate context is treated as suspect. Accounts are penalised for discussing abuse rather than promoting it. Moderation systems appear unable-or unwilling-to recognise context. Appeals are inconsistent, opaque, or practically impossible to navigate. Meanwhile, people who deliberately misuse reporting systems to target survivor advocates, charities, journalists, and campaigners often continue without consequence.

The result is that people become more cautious about what they say publicly, particularly in areas where the work already attracts hostility and organised harassment. It also sends a troubling message to survivors and those supporting them: that even speaking clearly about abuse can carry a penalty.

What needs to change

We are not asking social media companies to stop moderating harmful content. We are asking them to stop treating accurate language about abuse as though it were harmful in itself.

1. Stop penalising accurate language used to describe abuse in legitimate contexts

Content about abuse should not be suppressed, removed, or used as the basis for account penalties simply because it contains the correct terminology.

There must be a clear distinction between discussing abuse and promoting it. If a post is educational, journalistic, part of a public awareness campaign, research-based, survivor-focused, or produced as part of safeguarding and public-interest work, that context should matter.

2. Build moderation systems that can recognise context

A post condemning sexual violence is not the same as a post encouraging it. A campaign challenging harmful reporting is not the same as harassment. A safeguarding organisation sharing information about child sexual abuse is not posting abusive content.

It should be obvious. Current moderation outcomes suggest otherwise.

If automated systems cannot reliably distinguish between harmful content and public-interest speech about harm, they should not be making these decisions without meaningful human oversight.

3. Take malicious reporting seriously

Bad actors know how to use platform systems against the people they want to intimidate or silence. That includes survivor advocates, abuse organisations, journalists, campaigners, and public educators.

Platforms should invest far more in identifying false, vexatious, or coordinated reporting and in protecting users who are clearly being targeted in this way. Reporting tools should not become censorship tools for trolls.

4. Offer meaningful review by actual humans

When content is removed or an account is penalised for discussing abuse, there must be access to a genuine review by someone capable of understanding context.

An automated notice with little explanation and no meaningful appeal is not good enough, particularly when the account belongs to an organisation providing support, education, advocacy, or reporting in the public interest.

5. Work with people who understand this field

Policies on abuse-related content should not be developed in isolation. Platforms should be consulting organisations and professionals with expertise in abuse, safeguarding, coercive control, child protection, violence against women and girls, trauma-informed communication, journalism, and survivor support.

Too often, the current approach appears to rely on word detection rather than an understanding of how abuse is actually discussed, reported, or challenged.

6. Be transparent about how these decisions are made

If content about abuse is being restricted, people deserve to know why. Organisations should not have to reverse-engineer platform rules through trial and error, removing accurate language in the hope of avoiding penalties.

Clearer guidance would not solve every problem, but it would make moderation decisions easier to understand, challenge, and improve.

This is not a niche issue

This affects charities, campaign groups, journalists, educators, survivor advocates, researchers, safeguarding professionals, and the people who rely on them for information and support.

It also raises a broader question about the kind of public conversation social media platforms want to enable. If platforms cannot distinguish between abuse and speech about abuse, they are failing in one of the areas where context matters most.

We believe social media platforms can do better than this, and they should.

That starts with directing enforcement towards the people using your platforms to harass, threaten, intimidate, and silence others-not the organisations trying to educate the public, support survivors, improve understanding of abuse, and hold those responsible to account.

We would welcome the opportunity to discuss these issues directly and to work with platforms on a more effective approach. Until then, we will continue to raise concerns about moderation systems that make it harder to talk honestly about abuse while doing too little to address those who deliberately weaponise those systems against others.