As part of our commitment to put the Trust back into Trust and Safety, I aim to provide Blacksky’s community with knowledge of what has been happening within moderation by showing our data and by demonstrating how both off and online events involving Black folks have a direct impact on our workload to keep Blacksky a safe space for all within the diaspora.
This isn’t a report on violence; it’s a reflection of our community’s trust in our team to handle content that, in their opinion, violates our guidelines. It also serves as a tool to learn more about our moderation and how aligned our definitions of the labels are with what’s being reported by our community.
What I hope is learned from this reflection is that our Trust and Safety team tries to understand a lot of cultural differences by searching for information, reading related literature, and using our community as a resource. We often get very detailed reports, which increases the speed at which we can issue a label.
Trusting your peers
In the month of July, we will officially launch our peer moderation system, which is not only something that the community wanted but a step in the direction towards a culture of digital resistance. Peer Moderators will be regular members of the community except for one thing: they will have the ability to label content that goes against our community guidelines, moderation policies, or could harm the broader community on Blacksky.
In the last month, we’ve given them training on all of our policies as well as training on our code of conduct. We believe that being a Peer Mod should be about wanting to help people understand how their harm can impact the broader community and that labels shouldn’t be placed based on our own biases, but rather to serve our community. You can read more about our expectations of peer mods here.
What happened on your feed
In the month of June, there were some large-scale events that had a direct impact on moderation. We looked at our Trending Topics, compared them to news directly related to Black and other marginalized folks, and as a result, were able to make suggestions to upticks we saw in 465 labeled posts.
During the sentencing of Karmelo Anthony on June 10th, we noticed an uptick in the presence of white supremacists coming on Bluesky to harass Black folks with AI-generated material of Karmelo being in prison, sexually assaulted by inmates, and other upsetting scenarios.
12.7% (58 posts) of all reported content for this month has been labeled Synthetic media with a clear uptick after news of sentencing came online. This label is used to provide context for artificially generated content uploaded by users.
On June 2nd, Rick Chow was found not guilty of the murder of Cyrus Carmack-Belton. Black folks came online to talk about an Asian Store Boycott. This was met with anti-Black rhetoric and anti-Black harassment across social media platforms.
On June 8th, after a stabbing occurred in Ireland, white supremacists came to social media platforms to ignite what would become a hate riot, during which they burned down cars and buildings believed to be owned or lived in by migrants. On Bluesky, it led to an eruption of anti-Black xenophobia.
From our perspective, both events led to an uptick in both anti-Black harassment and anti-Black rhetoric by naming Black folks’ response as disproportionate or an overreaction.
37.2% (173 posts) of all reported content for this month has been labeled Anti-Black Harassment, and 20% (93 posts) Anti-Black Rhetoric or White Supremacy. Anti-Black Rhetoric or White Supremacy is used to label content that reinforces ideas of white supremacy or (internalized) anti-Blackness.
Here’s a breakdown of all the other labels we’ve used this month:
*data shown is based on total labels excl. bots and spam reports
About 15% of all reported content (68 posts) were labeled for misogynoir
About 9% of all reported content (44 posts) were labeled for Adult content
1% of all reported content (5 posts) were labeled for doxxing
1.2% of all reported content (6 posts) were labeled for violence
The results are in: Listen to Black women
The UK government has started a conversation about X’s involvement in maintaining a culture of hatred by— according to The Guardian— endorsing far-right agitators who called for protests and a furious response after the stabbing in Belfast, leading to the hate riot.
They also banned social media for children under the age of 16, following in the footsteps of Australia, which banned under-16-year-olds from social media in December 2025. Australia believes that the ban is needed to protect the mental health and well-being of children online by eliminating the risks of social media, such as cyberbullying, harmful content, and online predators.
What is, however, interesting, is that there is no direct action being taken against platforms that endorse and refuse to ban hateful content or cooperate with researchers towards finding solutions. Western governments have long spoken in warning, but have never actually threatened to remove themselves or impose national bans. They’ve ignored concerns made by citizens, professionals, researchers, and even their own advisory boards about the growing nature of hateful sentiment and content.
Black cis and trans women and femmes have always been the target of tremendous amounts of violence online and have long advised on methods to minimize its growth. Moya Bailey’s astute analysis and coining of the term misogynoir in 2008, to describe the intersection between racism and sexism towards Black women and gender-expansive people, has led to research like Amnesty International’s Troll Patrol report in 2018 and a report on Digital Misogynoir by award-winning online abuse charity organization Glitch UK in 2023.
Amnesty International’s Troll Patrol put forth that 1 out of 10 tweets mentioning Black women politicians and journalists was abusive and or problematic. They also found that Black women were 84% more likely than white women to be mentioned in highly problematic or abusive posts. Amnesty International urged Twitter in 2018 to publish data regarding the scale and nature of abuse on its platform, and the company refused to do so.
Glitch UK analyzed close to a million text-based posts on social media and concluded that out of 20% (⅕) were highly toxic in regards to misogynoir and misogyny. They also advised social media companies to release transparent data reports and to include clear definitions of and policies on misogynoir, misogyny, racism, and white supremacy, amongst other things.
And this isn’t even reflective of the bulk of research into online violence regarding marginalized communities since 2010. The results are clear: Governments have ignored Black women when they rang the alarm of harmful content online, and now, instead of more safety tools, users online have less access to proper resources when encountering it, to the point where social media has been rendered a hazard for the youth by those same governments.
Blacksky’s wish to put the Trust back into Trust and Safety started with Dr. KáLyn “Kay” Coghill’s desire to have transparent labels that directly address the harm occurring to Black women, femmes, gender-expansive folks, and Black people writ large. We aim to make more specific labels to accurately describe the types of online violence Black and other marginalized folks experience online. If you wish to read more about Misogynoir, Dr Kay wrote a leaflet on it that you can read here.
Thank you for reading, next month I will be diving into more patterns of harm reduction, transparent data, and hopefully our first successful month of Peer Moderation!