Glitch censorship
There is a particular kind of censorship that has kicked in the last couple of years that doesn’t announce itself like the good old fashioned Insta-jail. No warnings. No policy violations. No community standard reminders. Just a sudden irritating kind of friction that is confusing, deniable, and easy to dismiss as a “bug.” Maybe my phone is too old? Maybe I need to update my operating system? Maybe the Wi-Fi is iffy. Maybe, maybe, maybe.
It is the digital equivalent of those spikes nasty high street brands put in front of their doors to stop unhoused people from lying down. It’s a closed door that is open for everyone but YOU – and those like you, who share news on ecocide, genocide, or “Plasticine”. Don’t you even dare to share posts on F1lton twenty-four or m1litary budgets or surve1llance.
You refresh the page.
You wonder if it’s you.
Even worse- friendships are destroyed as people accuse you of unfriending them. It happens, not once, but twice or more.
This is glitch censorship: 2025 built-in suppression that lurks beneath the surface of platforms. It poses as a technical issue, feels like a reprimand, but is something far more heinous. It’s a silent two fingers. They know, you know, the movement knows, but say anything and you sound paranoid and delusional.
When did it start?
For me, looking back, I think it started in in the fall of 2023 with a mysterious “unfriending” of followers, after i received a certain video from someone in the BlueWhite nation. I put it down to genuine ‘unfollows’ because of the nature of the posts I was sharing. We are steeped in anti-”arab” sentiment after all. BIPOC men have long been dehumanised in western culture and still are. The propaganda was and still is intense.
The glitches became more frequent around early 2024 and then sporadic, only to return and become really noticeable at the beginning of this year. Glitches have appeared during specific times like an invisible brake that is pressed from time to time, purposefully, and far too specific to be random.
In January 2025, Mark Zuckerberg announced that Meta is eliminating its third-party fact-checking program in the United States and instead implementing an X-style “community notes” system where users add context, or report-censor- “potentially misleading posts.” Suddenly the brakes were off when it comes to insulting comments, but there was a cascade of automated reports coming from third party apps.
Since 2015, an agency within the Blue-White government has sent numerous content removal requests to social media platforms, including Meta (the parent company of Instagram). Meta has a high compliance rate with these official requests for content deemed to violate terms or laws. According to internal Meta data leaked by whistleblowers, Meta complied with 94% of takedown requests issued by the 1sraeli government since Autumn 2023. One whistle-blower revealed that over 90,000 posts were removed to comply with these requests in one action, often in an average of just 30 seconds. AI was employed to review and mass censor accounts and posts, rather than the standard practice of human moderators overseeing government requests. This is not speculation, it is out there in the public domain. In addition, the vast majority of criticism against the Blue-White state’s actions (95%) fell under Meta’s “t3rror1sm” or “v1olence and 1ncit3ment” categories. A shift that certain lobbying groups want to be embedded into law. In Sweden, every single complaint of anti-sem1t1sm is automatically investigated by a government body, so the system is easily abused by pro_iz bodies that have plenty of money to tie up universities and journalists in legal battles that the lobbyist fully expect to lose. The aim is friction, exhaustion, fear – and to drive people to self-censor.
Third party interference
There are many examples of Iz run efforts, aside from Meta compliance. A group called Act.IL ,for example, is an online campaign and app (now defunct since 2022) created by 1srael’s Ministry of Strategic Affairs to bolster pro-1srael support and counter the B0ycott, Divestment, and Sanctions (B D S) movement online. It provided “tools for Z10 activists” to easily respond to content deemed critical of 1srael. According to an IT colleague, these “activists” include paid data centres in India, NW Africa and the Middle East that also run fake Hasbara accounts mainly on X. You go on a website and put in an account and the account is targeted with bot-like actions to trigger platform censorship. Anyone can use these websites, but they are blocked in the Nordic region.
The “Iron Truth” is an 1sraeli tech volunteer group that launched soon after the 0ctober 2023. Members of this group used their personal connections with employees at major tech companies (including M3ta) to flag and facilitate the removal of content they considered harmful to 1zra3li interests. A Telegram bot was used to mass-report content, whilst a major 1srael1 IOF extreme hat3 account was not taken down for the longest time.
Screenshot from Haaretz. Iz media can sometimes be more open on offences than western media….
AI-Driven Campaigns by Stoic & Bridges Partners LLC were commissioned by The Blue-White “Ministry of Diaspora Affairs” to push covert influence AI campaigns to create fake accounts and posts on platforms like Instagram to generate pro-1srael sentiment, particularly targeting U.S. lawmakers. We’ve all seen the fake demonstration posts that appeared early on, with crowds of demonstrators with 8 fingers and architecture that didn’t align.
In addition, 1srael1 tech companies like Zencity, which emerged from the m1litary intell1gence Unit 8200, develop surve1llance technology used to mon1tor social media activity. Other firms like Cyabra specialise in identifying what they call “disinformation campaigns.”
1zra3l1 location-based feature apps like Reporty: dubbed “999 Instagram” in some news reports, are designed for video calls to emergency services, allowing dispatchers to visually assess a crisis rather than relying solely on verbal descriptions. This is for emergency use and not for general social media reporting but show how 1sra3li players are increasingly moving into the location based surveillance space with their own definitions of assessing a situation – or a user.
How is this data shared with 3rd parties? “Report It” is an app freely available in app stores for users to share details of antisem1tic incidents, including photos and videos, which are then forwarded to third party local organisations and law enforcement. This is problematic when the definition itself is being misappropriated and when social media platforms are freely sharing user data with oppressive governments or parties acting covertly in their interest.
X’s (Twitter) Location Tool: While not a BlueWhite app, is an X feature that publicly displays an account’s location. This was used by the 1sraeli Foreign Ministry to “expose” users claiming to be in Gza but posting from “other countries” to solicit donations. This was misleading information as many Gzans have 1sraeli mobile numbers – especially older phone mobile numbers. They are, after all, under total occupation.
Meta itself is clearly compliant. Human Rights Watch documented hundreds of cases where the mere neutral mention of the H word on Instagram and Facebook triggered the “DOI policy”, (Danger0us Organisations and Individuals policy) prompting the platforms to immediately remove posts, stories, comments, and videos, and restrict accounts that posted them. Once flagged, there is no information available on how long your account, or related accounts are listed and you are guilty by association.
While the DOI policy permits reporting on, neutrally discussing, or condemning designated organisations or individuals, it also states that “if a user’s intention is ambiguous or unclear” (to the algorithm) “we default to removing content.” So the default now is silencing. All cases that Human Rights Watch reviewed found that Meta removed even neutral mentions of the H word in relation to developments in Gaza.
This is also regional. Certain posts are more censored in the Nordic region: I have experienced both open censorship/ punishment and glitch censorship when sharing posts from the well-known account The Humanity Archive on racism. The account and posts are clearly critical of racism, yet I was penalised for sharing the posts whilst the original post and account stayed up. I have had the same experience when sharing NAT0 critical posts in Swedish and English, but not other languages.
Image: The anti-racist Humanity Archive has a large social media following, and yet my reposts were censored.
What exactly is glitch censorship?
Glitch censorship works alongside open censorship to ware us down when we are not in any direct violation of the community guidelines. It is the non-transparent throttling of content rather than outright bans. It is “soft warfare”; the endless spinning wheel, the story that never loads, the blank wall of someone’s Instagram account. Your account stops appearing in searches or repeat crashes. Your follower count drops, the engagement of “likes” and comments drops even further. A post refuses to upload. A link behaves normally on your end but shows up as “broken” for others. You post Taylor Swift to check if the ‘algo’ is really against you.
It is.
No messages of warning on the platform.
No one told you what line you crossed.
The ambiguity is part of the method.
They used to beg us to stay on their platform, now we are the “undesirables”.
Messages sent on Messenger fail to load.
These micro-obstacles are only placed in front of certain people, certain conversations, certain accounts, certain ideas. Going back to the start- it was in Autumn ‘23 when I was sent a (now debunked) atr0c1ty pr0paganda video on Messenger from someone I was loosely connected to in 1sra3l.
I watched the horrific video, and my FB account was then blocked within seconds of the video ending. I was shocked, upset, of course. That video itself deserves more attention than I can give here- so will stick to the point. Needless to say, her account wasn’t shut down. I got a warning, several warnings, but then I could not write any messages, even on my own posts, for over 3 weeks. Looking back, the glitches started after this event, primarily with mysteriously losing followers. As I had in the region of 4000 “friends” on Facebook at that time, there was no way of knowing if a connection had been broken on purpose unless they told me.
I did get some messages: “Why did you unfriend me? I like you!” was one message I got. I put up a post stating I might have been ”hacked” but that post went unseen. I was shadowbanned.
Heartbreakingly, another vague acquaintance wrote a very angry message accusing me of unfriending her because she is Jewish. I had already explained it was nothing to do with me but I wasn’t believed the second time the connection “broke”. I also reacted poorly, I was upset that she didn’t even bother to ask if I had unfriended her- or if I did – why. She was I felt, centring her feelings over the horror at we were witnessing on a daily basis. I decided to shut down my FB account to avoid further confusion.
I continued on Instagram, reluctant to leave and lose my contacts there. Some of us jointly decided to migrate our chats to Signal, but ironically, most direct news is from Instagram, despite the censorship.
Time went on, more and more people joining the calls for justice. The glitches happened particularly frequently the week when the journalist Anas al-Sharif and colleagues were murd3red in an air 1sra3l1 strike. It already feels like a lifetime ago, so much happens every day. Atrocities are buriied in more atrocities. The glitches increased with mutual aid fundraising posts from a surgical nurse, and in October 2025, Dr. Mu’ath Abu Rukbeh, a Palestin1an vet was k1dnapped and k1lled in northern G’za. Moaz Abo Rokba, one of the strip’s few vets who can perform surgery was also k1lled in an 1srael1 strike. Mountains of crimes against humanity, disbelief and grief.
And Gza is not the only target of the censorship. Many others are reporting similar unresolved digital interference incidents that mirrors broader patterns of algorithmic suppression. BIPOC people face selective moderation than their white counterparts.
I spoken to one Queer BIPOC influencer who has noticed “glitches”. While they agreed that I mentioned them I am reluctant. They wanted to emphasise tha they don’t feel Meta is targeting them personally they agree that the timing of glitches and selective removal of content, and parallel security breaches raise legitimate concerns about potential systemic or coordinated factors- institutional, algorithmic, or third-party.
Spend enough time on these platforms and you will start to notice the patterns:
– Followers mysteriously unfriended: Not a mass exodus, just a steady leak. A kind of social shadowing.
– Instagram posts that simply won’t upload: No error message, no explanation. Just persistent failure.
– Accounts temporarily “invisible”: Friends message you asking why you deactivated—except you didn’t.
– Broken links: They work perfectly when you test them, yet others receive only errors.
– Disappearing comments or DMs that never send.
– Links you send are broken for others
– Cannot share an Insta post to Facebook
– Posts downgraded in visibility, sometimes you can get a message that this has happening but usually not
– Previous posts mysteriously missing
Deniability is part of the abuse.
Is it possible any of these is a bug? Of course. Deniability is also part of the abuse. Platforms are enormous, fragile, living systems of code. Bugs happen, especially when code is being messed with on a daily basis. But the glitches are always against specific content about Pal3st1ne, so the message is received loud and clear. There is no effort made to even hide it. When the same friction appears only around topics of human rights violations, climate crisis, sharing Gza based accounts, mutual aid, critiques of power, etc. the patterns are blatantly intentional.
November 2025 at the annual conference of the Jewish Federations of North America in Washington, D.C. 1srael1 military spokesperson Dan1el Hagar1 states that !sra3el has lost the w*r on social media.
D’ya think???
Sarah Hurwitz, former chief speechwriter for Michelle Obama and a senior speechwriter for President Barack Obama, has called for Jewish schools to ban smartphones until senior year, but worse than that, she argues young Jewish people ‘can’t hear anything through the wall of de*ad children in Gaza.” Astonishing that anyone can say this so openly and so shamelessly. Hillary Clinton recently said that misinformation on social media has shaped young people’s views on Israel and Palestine in a dangerous way – meaning not a pro-Iz way.
The fact that they are lamenting young people’s reactions to witnessing genoc!de, and not screaming from the rooftops for the genoc1de to stop is completely astonishing.
The extreme violence continues but there are fewer violent images- so many local journalists are no longer with us, and foreign journalists are blocked from entering – but it is more than that.
There is a silent ongoing censorship that continues.
Just silent, persistent, cowardly glitch censorship.
Glitch censorship is yet another attempt to make life harder, slower, more and more discouraging. And they want to push us to other platforms, to remain in bubbles preaching to the converted. Glitch censorship is just yet another narcissistic abuse method, deniable by design. It is tech “härska teknik” It’s gaslighting. You cannot possibly prove it. There is nowhere to complain to- if you do you are complaining to the culprit.
Don’t doubt your perceptions
I believe that employees know about glitch censorship, and I hope more whistleblowers will step forward. We’ve seen demonstrations at annual meetings, and internal conferences. But mostly we don’t complain because there is nowhere to complain to. And so the suppression continues unchallenged, undocumented, and unnoticed, along with the broader media landscape complicit in manufacturing support for gen0cide, which is, by the way, a prosecutable offence at the Hague. We are silent. After all, it’s nothing compared to the actual atroc1ties happening on a daily basis. Being connected to people I care about outweighs my desire to boycott. The platform gets to maintain its illusion of being the democratic domain, without ever admitting their hands are on the scale. You get minimum functionality that keeps you there. Enshitification is the new term. When censorship masquerades as malfunction, it gaslights the user into doubting their own perception, postponing action. It’s a series of micro-aggressions designed to wear you out and wear down your sense of agency.
They know we are worn out, desperately trying to grapple together funds so that one family can buy a tent, another food or medicine. They know we have to conserve our energy. It is easier to exhaust us than silence us.
What can we do about it?
In the long run we need to start migrating off these platforms. We need to nurture AFSM (Away from Social Media) relationships, if not AFK (Away from Keyboard). There is no simple solution. Content creators rely on incomes from social media, including YouTube that is also censoring BIPOC and Palest1inian creators disproportionately. These platforms are directly working against all of our causes and complicit in whitewashing and manufacturing consent for g3n0cide and occupation – from Gza and West Bank to Congo, Sudan and beyond.
Glitch censorship is part of the logic of Empire.
But it shows desperation.
Awareness helps. And so does documentation. Screenshots, screen recordings, timestamps, these small acts of witness turn glitch censorship from an ineffable suspicion into something trackable. Keep receipts as there will be a reckoning. The code owners have names. There will be whistleblowers. Accountability takes time and effort. When individuals gather notes, patterns are documented for future court cases. When communities speak up, silence becomes harder to enforce.
Support organisations seeking accountability.
The H1nd Raab Foundation is one to donate to.
Do not self-censor. Recording important videos on your phone by recording the screen (see settings). Share that. Very often it is the original account that is shadow-banned and not the video. Find it on another platform and download it. Tik Tok is good for that although that’s a whole other story.
We need to support ethical tech – technology designed, developed, and used in ways that prioritise human well-being, fairness, privacy, and societal good. It focuses on minimising harm, avoiding bias, ensuring transparency, and respecting users’ rights. Ethical tech encourages companies and creators to consider the long-term social impact of their innovations rather than just profit or efficiency.
Open-source tech is technology whose source code is freely available for anyone to view, use, modify, and share. It promotes collaboration, innovation, and community-driven improvement. Open-source projects are often more transparent and trustworthy because the code can be openly inspected and audited. Together, ethical tech and open-source tech support a future where digital tools are built responsibly, inclusively, and with shared benefits for society. We need mass-FOSS social media.
These privately-owned platforms owned by the “broligarchy” means our freedom of expression is fragile, even when it appears intact.
We have to spread awareness and de-centre our lives off these corporate platforms and move into open-source space. It is time to switch to ethical tech. To FOSS. Free and Open-Source Software. To use and donate to those initiatives and eventually cut the cord.
Because censorship these days rarely looks like censorship.
It looks like buffering.
It looks the quiet removal of content
It looks like a missing or greyed-out button, missing notification, an unloaded message or a broken link.
It will look like a glitch – or it will not be visible at all.
Unless we name it and challenge it, we cannot understand it.
We are not imagining it.