Ofcom, the UK's media regulator, formally warned online platforms in August 2024 that they risked becoming tools to stir up hatred. The warning came directly after violent disorder swept across multiple English towns and cities — and the regulator had found evidence that social media posts had helped fuel what happened on the streets.

What Sparked the Unrest

Between 30 July and 7 August 2024, anti-immigration riots spread across the United Kingdom, with some targeting mosques and hotels housing asylum seekers, according to parliamentary records. What caught regulators' attention was how quickly and widely the disorder spread. This was not unrest confined to one town. It was dispersed, fast-moving, and coordinated partly through social media.

BBC reporting from October 2024 confirmed that Ofcom found a direct link between posts on social media and the violence that followed. Social media was not simply recording what was happening — it was actively helping to spread and organise the disorder.

How Regulators Responded

Ofcom sent an open letter to online platforms on its behalf, but it stopped short of issuing formal punishments. Instead, it made clear expectations: platforms should check their systems against their existing legal duties to block content that incites racial or religious hatred.

The UK government moved quickly to say this was the right approach. Government statements published on 8 August 2024 said Ofcom had signalled clearly that social media companies needed to act. This framing mattered because the Online Safety Act — the law that gives Ofcom these powers — had only just come into full effect. An open letter created a documented record and shifted responsibility toward the platforms, even if new laws had not yet been fully tested.

Why Social Media Spread the Riots So Fast

Analysis published by Tech Policy Press on 21 August 2024 noted that social media platforms did more than just share information. Their algorithms — the software rules that decide what to show users — actively amplified inflammatory posts and sent them to people most likely to engage with them.

Think of it this way: in the past, spreading a message across multiple cities took time. Someone had to physically travel or use slow communication. With social media, the same message can reach thousands of people across the country in minutes. Disorder that might have been contained to one place in earlier decades jumped between cities within hours.

What made this different from earlier concerns about online radicalisation was the speed and the scale. The coordination layer was a handful of platforms — mainly Facebook, X, TikTok — operating under rules designed for a slower-moving information environment. The systems governing the old broadcast era — where a handful of editors decided what millions saw — could not keep pace with a system where anyone could publish to thousands instantly.

The Broader Problem

Ofcom had already been working on protecting children from violent and hateful content online. Proposed measures published in May 2024 — before the July riots — included plans to shield young people from abuse and hateful material. The riots provided a real-world test of the problem those proposals aimed to solve.

Ofcom's Adults Media Lives 2025 study later found that ordinary people had experienced a noticeable surge in hateful and violent content tied to the riots. This was not a theoretical concern — it was something everyday internet users actually encountered.

What Regulators Can and Cannot Do

The Online Safety Act gives Ofcom significant powers: it can require platforms to carry out safety audits, publish transparency reports, enforce codes of practice, and impose large fines. But there is one thing it cannot do — at least not in the heat of a crisis. It cannot step in and remove content in real time. That job remains entirely with the platforms themselves and their moderation teams.

This gap between what regulators can do and the pace at which harmful content spreads is not a flaw in the law's design so much as a structural fact. No regulator can become a live content moderator for the entire internet. The real test for the Online Safety Act's next phase is whether it can make platforms' systems resilient enough to slow down harmful amplification before unrest reaches the streets in the first place.

Where This Leaves Us

The parliamentary inquiry that followed the riots shows that lawmakers remain committed to tighter accountability for social media companies. These platforms still operate with recommendation systems that tend to amplify high-emotion content — outrage, fear, messages that make people feel a strong tribal connection — regardless of whether that content breaks the platform's own rules. Fixing that deeper mismatch between what the rules say and how the algorithms actually work is the longer-term challenge.

What this moment offers is something concrete: real UK evidence from an actual crisis that regulators and lawmakers can use to test whether the Online Safety Act is working, or whether it needs to be strengthened. That evidence, built during an actual emergency rather than in theory, will shape how these platforms are policed for years to come.