You don’t usually see lawmakers open a public help desk for a law they just passed. But that’s exactly what’s happening in the UK. Parliament is currently accepting submissions for evidence regarding the effects of the Online Safety Act. They want to know about its successes, sure, but more importantly, they are asking for its failures and detrimental effects. The deadline is September 2026.
Let’s be clear about what this means. The government passed a massive piece of digital legislation, and almost immediately realized it might be a disaster. If you use the internet—which is everyone reading this—this Act is currently dictating what you can say, see, and share. And the people who wrote it are now asking, ‘Hey, did we break things?’
Most of the debate around this law has been a tired, binary shouting match. On one side, people scream ‘Censorship!’ On the other, people scream ‘Protect the children!’ It’s an exhausting cycle that completely misses the actual battlefield. The real threat isn’t the content being taken down; it’s the machinery pushing it in the first place.
You can’t fix a poisoned river by arresting the people drinking the water; you have to stop whoever is dumping the toxins upstream.
The Online Safety Act obsesses over takedown rules. It demands platforms remove illegal or harmful content. But it fundamentally fails to address algorithmic accountability. When a platform complies with the Act by removing a harmful post, the job is considered done. Yet, the platform’s recommendation engine—the invisible hand deciding what goes viral—continues to learn from that interaction. It continues to push the next user down a radicalizing rabbit hole.
We are treating the symptoms while subsidizing the disease. Platforms aren’t just passive bulletin boards where bad actors happen to post; their ad-driven business models actively amplify outrage, polarization, and harm because that is what generates engagement.
Regulation designed to protect the vulnerable is a placebo if the platform’s business model relies on exploiting them.
This call for evidence is a rare moment of regulatory humility, but it’s also a flashing red siren. The damage from poorly designed regulation is already compounding. Every day this law operates in its current state, it hardens into precedent. It sets the legal DNA not just for online speech in the UK, but for global norms. If the UK gets this wrong, other governments will copy the flawed blueprint.
If you’ve ever felt like your feed is actively trying to make you angry, you aren’t imagining it. The same mechanisms that promise to keep you safe are quietly weaponized to keep you scrolling.
We have until 2026 to course-correct. If we waste this evidence stage arguing about basic censorship instead of demanding algorithmic transparency, we lose. The platforms will happily comply with takedown notices while their engines continue to silently dictate our digital reality. The law isn’t just deciding what gets deleted; it’s deciding what gets amplified.
FAQ
Q: Why is Parliament asking for evidence on a law they just passed?
A: Because they realize the unintended consequences are already compounding. It's a rare admission of regulatory humility, acknowledging that top-down rules on digital speech often break upon contact with reality.
Q: Why does focusing on takedown rules miss the point?
A: Taking down a harmful post doesn't stop the algorithm from having amplified it in the first place, nor does it stop the recommendation engine from pushing similar content to others. It treats the symptom, not the disease.
Q: What is the actual solution to online harm?
A: Algorithmic accountability. We must force platforms to be transparent about how their recommendation engines amplify outrage and harm, rather than just playing an endless game of digital whack-a-mole with individual posts.