The timing is so perfect it almost looks scripted. On June 8, 2026, a 30-year-old Sudanese asylum seeker named Hadi Alodid stabbed a man named Stephen Ogilvie on Belfast’s Shankill Road, slashing his eyes, face, and back in what witnesses and police described as an attempted beheading.
Ogilvie, 44, is in hospital. By the following evening, the Northern Ireland city was on fire.
Cars, houses, and a bus were torched. Masked protesters kicked in doors along the Lower Newtownards Road. Riots spread to Portadown, Antrim, and more. Protests erupted in Glasgow, Edinburgh, Southampton, and outside Parliament in London.
And on that very same day, June 9, Britain’s speech regulator Ofcom published its finalized “crisis response protocol,” a new set of rules pressuring every major platform in the country to build censorship infrastructure that activates whenever someone at Ofcom or in government decides a “crisis” is happening or about to happen.
Reclaim Your Digital Freedom.
Get unfiltered coverage of surveillance, censorship, and the technology threatening your civil liberties.
The protocol was written because of the 2024 Southport riots. If you wanted a clearer illustration of how these powers will be used, you couldn’t commission one.
The Official Response Tells You Everything
Within hours of the Belfast stabbing, the political class was already running the Southport playbook.
PSNI Chief Constable Jon Boutcher warned: “The challenge we face with today’s online toxic nature is that people are incited by people who are faceless and know nothing about this brilliant, vibrant place. Do not be fooled or duped by people online.”
First Minister Michelle O’Neill said: “of those people out there who are stoking up tensions in that social media space who are happy to raise tensions, they do not represent us. We are good people and I don’t want to see anybody living in fear.” Justice Minister Naomi Long blamed “bad faith actors” online for stoking tension.
UK Prime Minister Keir Starmer called the stabbing “sickening” and made clear there was “no tolerance” for street violence.
Notice the pattern. The question of why people are angry gets replaced by the question of who shared the video. The people who are upset about the attack online become the problem and a censorship target.
And the regulator that just published new tools for suppressing online speech during exactly this kind of event gets to look prescient instead of opportunistic.
The script never changes. The censorship methods, though, keep growing.
What the Protocol Actually Does
Ofcom’s crisis response protocol, added to its Codes of Practice under the Online Safety Act (OSA), pressures social media platforms, messaging apps, forums, dating services, and essentially any website where humans can talk to other humans to build standing censorship systems for “crisis” situations.
The OSA’s Codes aren’t technically mandates. Platforms can theoretically ignore them and “use other effective measures to protect users from illegal content and activity.” This is like saying you don’t have to wear a suit to court. Technically true; inadvisable if you plan on winning.
Under the protocol, covered platforms must build systems to identify and respond to spikes in “illegal” content and “content harmful to children” during what Ofcom defines as an “extraordinary situation in which there is a serious threat to public safety in the United Kingdom.”
They should deploy a temporary crisis response team “as soon as reasonably practicable if the provider determines that a crisis is occurring or is likely to occur.” They should run post-crisis reviews to evaluate how effective the censorship was. And large platforms should open a dedicated communication channel with law enforcement, so police can phone in their content complaints on a priority line during the crisis.
A direct hotline between police and a platform’s moderation desk, activated during civil unrest, is a state-to-delete pipeline. There’s no court order, just a cop calling a content moderator and saying “this one,” and the post vanishes.
“Likely to Occur” Is a Loaded Gun Pointed at Every Trending Topic
The most dangerous word in the entire framework is “likely.” Platforms don’t need an actual crisis and they don’t even need burning cars. They need a feeling that trouble might be on the way.
Ofcom expects them to activate the protocol when a crisis is “likely to occur,” which is a subjective, future-tense judgment call being outsourced to private companies with every financial incentive to over-comply.
Think about what this means in the context of Belfast. A video of a stabbing goes viral. People start posting angry reactions. Some of those reactions call for protests. Under this protocol, a platform looking at rising engagement around the Belfast attack could decide that a crisis is “likely,” activate its response team, and start suppressing content before a single car has been set on fire.
The censorship arrives before the crisis. By the time anyone asks whether the deleted posts were actually illegal, the moment has passed.
And the definition of “crisis” is elastic enough to be meaningless. Ofcom says an “extraordinary situation” can include “local or regional crises” and even “an overseas event.” A protest in one neighborhood. A political incident in another country. Something trending on X that makes a regulator nervous. Ofcom gave itself a rubber stamp that reads “CRISIS” and left the ink bottle open.
Who Gets Caught?
Three categories of “user-to-user services” are expected to comply: large platforms rated high-risk for “terrorism, hate, harassment, stalking, threats and abuse, and foreign interference;” large platforms “likely to be accessed by children” rated medium-risk for “abuse, hate, and violent content;” and any platform of any size “likely to be accessed by children” rated high-risk for those categories.
The OSA covers basically any website where one human can communicate with another. Social media, messaging apps, forums, dating apps, video and photo sharing, online marketplaces, multiplayer games, and AI services whose output can be shared. The exemptions are email, SMS, MMS, and one-to-one voice calls. So the technologies of 1995 are exempt. Everything invented since is in scope.
If you wanted to design a system where the government could control what people say during the exact moments when free speech is most essential, you’d build something very close to what Ofcom just published. And you’d probably have the nerve to call it a safety measure, too.
