When a terrible event happens, the first hours online can decide whether a community stays calm or whether rumours spill into the street. From 2 November 2026, large platforms covered by the Online Safety Act are expected to have a written plan for those hours and a team ready to act on it.
In this article we will discuss what Ofcom’s crisis response measures say, which services they reach, how they came about, what critics fear, and what families and small site owners can do now.
The Measures That Took Effect on 2 November
According to TechTimes, which is the main source for the detail in this article, Ofcom’s crisis response protocol measures began to apply on 2 November 2026. The report’s headline describes the start as a Sunday, which does not line up with that date, so treat 2 November as the date given in the body of the report. Ofcom signed the amendments to the relevant codes of practice on 12 October 2026, after finalising its decision on 9 June 2026.
These measures are not a free-standing law. TechTimes reports that they are recommended measures inside two of Ofcom’s codes of practice under the Online Safety Act: the Illegal Content Codes of Practice for user-to-user services, and the Protection of Children Code. A code of practice is guidance that sets out steps a service can take to meet its legal duties. A platform is not forced to follow the recommended steps word for word, but, as explained later, following them brings a significant practical advantage in showing that the duty has been met.
The practical effect is a change of expectations. Until now the codes mainly asked platforms to deal with illegal content and risks to children as part of everyday moderation. The new measures add a specific layer for extraordinary moments, when a sudden surge of harmful posts could overwhelm ordinary processes. They ask in-scope services to plan before a crisis, to act quickly during one, and to learn from it afterwards.
Because the information comes largely from a single secondary source, it is sensible to read the codes themselves on Ofcom’s website if you run a service or advise one. Details such as thresholds and definitions can matter a great deal, and a news summary cannot carry every qualification.
How the Rules Came About
The rules have a clear starting point in the Southport attack in July 2024, in which three girls were killed. TechTimes reports that false claims on social media about the attacker’s identity helped to fuel disorder across the country in the days that followed. That episode showed how quickly an untrue story can travel, and how difficult it can be for a platform to react when a flood of posts arrives faster than any team can read them.
Ofcom began work on the crisis response measures on 23 October 2024, according to the same report. It finalised its decision on 9 June 2026, and the code amendments were signed on 12 October 2026 before taking effect on 2 November. That is a little over two years from the start of the work to the start of the rules, which is typical for regulation that involves consultation, evidence from industry and legal drafting.
It helps to place this within the wider Online Safety Act regime. The Act gives Ofcom powers to set codes, to require information from services, and to enforce against those that fail their duties. The crisis measures are one more piece in that framework, designed for a narrower, sharper problem than day-to-day moderation. They do not create new categories of banned content, and they do not turn platforms into publishers of record. They are about preparation and speed.
Which Services Fall in Scope

Not every website or app has to follow the new steps, and size alone does not settle the question. TechTimes reports a two-part test, which applies in the same way to the illegal content code and to the children’s code, for services likely to be accessed by children.
The Size and Risk Thresholds
The first group is large user-to-user services that have been assessed as medium risk or above for hosting priority illegal content. The report gives terrorism, hate speech, harassment and threats as examples of the kinds of content involved. The second group is any service of any size that has been assessed as high risk. In other words, a small forum that carries a high risk of priority illegal content can be caught, while a large service with a low assessed risk may not be.
This is why the risk assessment matters so much. Under the Act, services must assess the risk that illegal content appears on their platform, and the result of that assessment feeds into which of Ofcom’s recommended measures apply. A site owner who has not completed one has no reliable way of knowing which side of the line they fall on.
Who Is Excluded, and What About AI Tools
TechTimes reports that search services are excluded from these measures. Generative AI services are not excluded as a category: according to the report, they can fall in scope if they operate as user-to-user platforms and meet the size and risk criteria. A chatbot that lets users share content with each other, for instance, could be treated differently from one that only answers a single user.
The table below summarises the scope test as reported.
| Type of service | In scope of the crisis measures? | Condition reported |
| Large user-to-user service | Yes | Assessed at medium risk or above for priority illegal content |
| User-to-user service of any size | Yes | Assessed at high risk |
| Service likely to be accessed by children | Yes | Same size and risk thresholds, under the children’s code |
| Search service | No | Excluded |
| Generative AI service | Possibly | Only if it works as a user-to-user platform and meets the size and risk criteria |
Defining a Crisis
The word crisis could stretch to almost anything, so the definition is crucial. TechTimes reports that Ofcom defines it as “an extraordinary situation in which there is a serious threat to public safety in the United Kingdom”. The threat can be local, or it can arise overseas, provided it affects public safety in the UK.
To picture how the definition might work, consider an illustration with invented details, not a real case. Imagine a violent incident in a British town in the early evening. Within an hour, posts appear naming a person as responsible, some with edited images, and thousands of accounts begin sharing them with calls for people to gather.
A platform with a protocol would want its indicators to flag that pattern: a sudden rise in reports, a cluster of posts using the same name, and contact from agencies. The senior team would then decide whether the situation is a crisis, and, if so, switch on the surge measures. The same platform might see a heated argument about a political announcement, which is intense but not a serious threat to public safety, and would not treat it as a crisis.
The report also lists what is left out: environmental disasters, pandemics and misinformation campaigns are not treated as crises for these purposes. That exclusion is easy to misread. It does not mean platforms can ignore false information during an attack. It means that a misinformation campaign, on its own, is not what triggers the crisis protocol. A violent incident that is followed by a wave of false claims might well be one.
Who Decides That a Crisis Is Under Way
According to TechTimes, the platform itself decides whether a crisis is occurring or likely. A notice from the Secretary of State under section 175 of the Act is an indicator, but it is not a determining one. That puts responsibility, and judgment, in the hands of the service, which has the live view of its own systems. It also leaves room for disagreement afterwards about whether a service reacted too early, too late or not at all.
Smaller services should note that the measures are scaled by risk and size, so the effort expected of a lightly used site is not meant to match that of a global network. Because the post-crisis analysis can be requested by Ofcom, the platform’s reasoning is likely to be examined later. A sound approach is to write down, in the protocol, the indicators that would lead the service to conclude that a crisis is likely, so the later explanation does not rest on memory.
Platforms Are Asked to Prepare, Respond and Review

The core of the measures is a cycle of four steps and one standing channel. TechTimes sets them out as follows, and the description below attributes each element to that report.
An Internal Crisis Protocol
First, in-scope platforms are expected to write an internal crisis protocol. It should list the indicators the service will monitor, which might include a sudden jump in reports, a spike in posts using certain terms, or signals from outside agencies. It should also set out how those indicators will be reviewed in real time. A plan that sits in a drawer and is checked once a year would not meet the spirit of this, because the point is to catch an emerging situation while it is still manageable.
A Senior Cross-Functional Team
Second, platforms are expected to keep a cross-functional crisis response team that is senior enough to make enforcement decisions without delay. The detail is important. A team of junior moderators with no authority to remove content or change settings would struggle in the middle of a fast-moving event, because every decision would need to be escalated. Cross-functional here suggests a mix of trust and safety, legal, policy, engineering and communications staff, though the report does not list exact roles.
Deployment and Surge Handling
Third, the team should be deployed as soon as reasonably practicable once the platform judges that a crisis is occurring or likely, and the platform should activate systems that handle surges in harmful content. In plain terms, that could include additional moderation capacity, tighter limits on resharing, or faster routes for reviewing reports. The measures leave the specific tools to the service, which is sensible, as a messaging-style app and a video platform face different technical problems.
Post-Crisis Analysis Within 90 Days
Fourth, platforms are expected to complete a documented post-crisis analysis within 90 days of a crisis ending. If the crisis is still going, the 90 days run from its start. Ofcom can ask to see this analysis. The idea is to build institutional learning, so that the next response is better than the last, and to give the regulator a record that can be tested.
A Dedicated Channel for Police and Agencies
Finally, TechTimes reports that platforms with more than 700,000 UK users must keep a dedicated channel through which the police and other agencies can share crisis-related information. Ofcom describes it as a coordination mechanism, not a replacement for court orders or legal information notices. Small and mid-sized services below that user figure are not required to keep the channel, though they can of course respond to lawful requests in the normal way. That limit is worth stressing. The channel is meant to help agencies and platforms exchange information quickly, but it does not give officers a back door to demand user data or removal of content without the usual legal process.
The next table sets the elements out side by side.
| Element | What is expected | Applies to |
| Crisis protocol | Written plan listing indicators and real-time review | In-scope services |
| Response team | Cross-functional and senior enough to decide without delay | In-scope services |
| Deployment | As soon as reasonably practicable once a crisis is judged to be occurring or likely | In-scope services |
| Post-crisis analysis | Documented within 90 days of a crisis ending, or of its start if ongoing | In-scope services |
| Agency channel | Dedicated route for police and other agencies | Services with more than 700,000 UK users |
Compliance, Safe Harbour and Enforcement
The measures are recommendations, not commands, but the Act gives them real force through what is often called a safe harbour. TechTimes reports that platforms which follow the recommended measures are treated as complying with the underlying safety duty under section 49. A platform that takes a different route must prove that its approach delivers an equivalent outcome.
For a large company with lawyers and a trust and safety department, the choice may be a close call. Following the recommended steps is the safer route, because it avoids an argument with the regulator about whether an alternative was good enough. A different design might fit a service’s technology better, but the burden of proof sits on the platform if Ofcom challenges it.
Documentation is the quiet theme running through the measures. A written protocol, a record of who sat on the response team, a log of decisions during an event and the post-crisis analysis together form a paper trail. If a regulator later asks why a platform acted when it did, that trail is the evidence. A service that did sensible things but kept no record will find it hard to prove them, which is why compliance staff tend to treat note-taking during an incident as part of the response and not an afterthought.
What the Enforcement Record Shows so Far
The report offers some context on how Ofcom has used its powers. It says Ofcom has issued more than £7 million in fines against 11 providers since October 2025. It also reports a £520,000 decision against 4chan, of which £450,000 related to failures to protect children. Separately, the report says Ofcom opened an investigation into X on 12 January 2026, narrowed it on 14 October 2026, and that it remains open. An investigation is a process, not a finding, and readers should not assume any outcome from the fact that one is under way.
These figures come from a single secondary report, and the report itself is cautious about parts of its own detail, so they should be treated as reported rather than as confirmed totals. The wider point they illustrate is simple: the regulator has shown it will use its enforcement tools, and services in scope of the crisis measures will know that.
Penalties for serious breaches of the Act can be large. The Act allows fines up to a fixed ceiling or a share of qualifying worldwide revenue, whichever is higher, so the exposure for a very large service is far greater than for a small one. Ofcom’s own guidance is the place to check how it intends to use that power in practice.
Critics Warn of Over-Removal
Not everyone is comfortable with the approach. TechTimes reports that the Open Rights Group and other civil liberties groups warn that removals made in crisis mode could sweep up lawful political speech. Their concern is that a platform acting fast, under regulatory pressure, may lean towards taking posts down when it is unsure, and that the people most affected could be those commenting on contested events.
The Speed Dilemma
The report also quotes the law firm Fieldfisher as warning that acting too slowly brings regulatory and reputational exposure, while acting too quickly risks restricting legitimate content. That dilemma is the heart of the matter. A platform that waits to verify every claim may be criticised for letting harm spread. A platform that takes down everything that looks risky may be criticised for silencing legitimate debate. No protocol can remove the trade-off, though a well-designed one can make the decision more consistent and easier to explain later.
The Comparison With the European Approach
TechTimes also reports that some critics say the UK scope is narrower than the crisis framework in the European Union’s Digital Services Act. A narrower scope may please those who worry about overreach and disappoint those who want broader protection. The report presents this as criticism rather than settled fact, and a reader who wants to judge it would need to compare the two regimes directly.
For readers trying to form a view, the balance can be put simply. Supporters would say the measures prepare platforms for the rare occasions when public safety is at stake and require them to act quickly and transparently. Critics would say the definition, the platform’s own judgment, and the pressure to act can combine to chill speech. Both positions can be held in good faith, and the evidence of how the measures work will only emerge over time.
Steps for Families and Small Platform Owners
Most people will never see a crisis protocol being activated, but everyone uses platforms during breaking news. A few habits help, whether or not a platform is in scope.
For Families and Ordinary Users
When something alarming breaks, slow down before you share. Look for the original source of a claim, such as an official statement from the police or a named news organisation, rather than a screenshot or a post with no origin. Be sceptical of posts that name a suspect, claim to know a motive or demand instant outrage, as these are the types of claim that spread quickly and are often wrong. If you cannot trace a claim, do not pass it on.
Use the reporting tools. Most platforms let you report a post for illegal content, harassment or threats, and you can block or mute accounts that keep posting abuse. Parents can check the age and privacy settings on their children’s accounts, talk about how rumours spread, and agree what the child should do if they see something frightening. If you see violent or illegal content, report it to the platform. If someone is in danger, contact the emergency services rather than relying on a report button.
Teenagers need extra care in these moments, because they often meet news first through short videos and group chats where context is stripped away. A calm conversation beforehand works better than a ban imposed during an event: agree that it is fine to ask you about anything alarming, that they will not forward a claim until it has been checked, and that they can leave a group chat that is spreading hatred without being thought rude. Keep in mind that the protection offered by the children’s code is a backstop and not a substitute for talking.
Media literacy is a skill that improves with practice. Ask who is saying something, how they could know it, whether anyone credible confirms it, and when it was posted. Old images and videos are often recycled during fast-moving events, and a quick search can reveal that a picture is from a different time or place.
For Small Platforms and Forum Owners
If you run a community forum, a gaming chat, a local group or a small app where users post to each other, start with the risk assessment the Act requires and keep it up to date. It will tell you how your service is likely to be classed, and TechTimes reports that a service of any size assessed as high risk can fall within the crisis measures. Read the relevant Ofcom guidance and codes directly, as this article is general information and not legal advice.
Even if you are not in scope, a short written plan costs little, and it will make any later conversation with a regulator, a host or your own users much easier. Decide who has authority to remove content or lock a thread, how you will receive urgent reports, how to reach the people who can help, and how you will record what you did. Tell your moderators what to do if a serious event occurs, and think through whether you could temporarily limit posting or resharing. A small volunteer team cannot match a global company, but it can be clear about who decides what.
Keeping the Trade-Offs in Mind
Whichever role you play, remember that safety and free expression pull in different directions. A cautious approach protects people from the worst content, and an open approach protects legitimate debate. The rules do not remove that tension, and the sensible response is to be transparent about decisions, to review them, and to correct mistakes quickly. If you are unsure whether your service is caught, ask a qualified adviser or contact Ofcom.
Frequently Asked Questions
When Did Ofcom’s Crisis Response Measures Take Effect?
According to TechTimes, they took effect on 2 November 2026. Ofcom signed the code amendments on 12 October 2026 after finalising its decision in June.
Which Platforms Have to Follow the Crisis Response Measures?
TechTimes reports that they cover large user-to-user services assessed at medium risk or above for priority illegal content, and services of any size assessed as high risk. Search services are excluded, and the same thresholds apply under the children’s code.
How Does Ofcom Define a Crisis?
As reported, it is an extraordinary situation involving a serious threat to public safety in the UK, whether at home or overseas. Environmental disasters, pandemics and misinformation campaigns are excluded.
Can the Police Use the Dedicated Channel to Demand User Data?
No. TechTimes reports that Ofcom describes the channel as a coordination mechanism, not a substitute for court orders or legal information notices. Services with more than 700,000 UK users must keep it.
Does Following the Measures Protect a Platform From Enforcement?
Following the recommended measures means a platform is treated as complying with the underlying duty under section 49, as reported. A platform that takes a different approach must show it achieves an equivalent outcome.
Should I Report False Rumours During a Major Incident?
Yes, you can report posts that are harmful or illegal using the platform’s reporting tools, and avoid resharing unverified claims. If anyone is in immediate danger, contact the police or emergency services directly.
