Sleepwalking into censorship: a reply to Nadine Dorries

In this week’s magazine I look at the threats posed by the so-called Online Safety Bill now making its way through the House of Commons. It gives sweeping censorship powers by creating a new category of speech that must be censored: ‘legal but harmful’. The government will ask social media companies to do the censoring – and threaten them if they do not. The idea is for the UK to fine them up to 10 per cent of global revenue (ie: billions) if they publish ‘harmful’ content – but harmful is not really defined. So censorship potential is wide open.

Nadine Dorries, the Culture Secretary, has suggested that the jokes of the comedian Jimmy Carr would be “harmful”. She has also suggested that Twitter ‘pile-ons’ would come under the remit of her fuzzily-defined law. This matters because no one really has a clue how the government will define ‘harmful’ – with so much cash at stake, Twitter, Google, Facebook and YouTube cannot afford to take any risks. They make almost no money publishing political content, yet stand to lose billions by getting it wrong. If in doubt, they will censor or shadow-ban – as never before.

Ms Dorries has taken exception to my article and said on Twitter (where else?) that it ‘is wrong in four key ways’. Rather than respond on Twitter, I’d like to take each of her points in turn here as it’s an important debate.

CLAIM
ONE: ‘Nothing in the bill requires companies to remove legal free speech. Our aim is to make tech firms protect their users and kids and uphold their own T&Cs’.

On the contrary, the whole point of her bill is to remove legal free speech. It would so so by creating a new definition: ‘legal but harmful’ (explained here). The definition of ‘harmful’ is staggeringly loose, making this Bill a censor’s charter. More importantly, her bill will inspire vastly more censorship than it will “require”. What happens – indeed, what is already happening – is that social media firms guess what might come close to being seen by the censors as ‘harmful’ and remove or shadow-ban accordingly. This is not a conspiracy theory or a fear but a statement of what is already happening. Examples:

• David Davis’s speech against vaccine passports removed by YouTube

• Socialist Workers Party page removed by Facebook

• Novara Media taken down by YouTube

• Professor Carl Heneghan and Tom Jefferson’s analysis of face mask data labelled ‘false information’ by Facebook.

Most popular

Ross Clark

Britain is living beyond its means

These are big ones that we know about (most were reversed after uproar) but there will be thousands of other shadow-bans and content removal that we do not know about. It’s often invisible even to the author, but I know about it as editor because we follow the figures and try to complain to the Big Tech firms (who, in Facebook’s case, ignore us). Expect much more of such unexplained censorship when her rules come into force.

2.
CLAIM TWO: “Right now tech bosses can arbitrarily remove posts even if they don’t break their T&Cs. We’ve seen countless examples of it happening. Our bill will end that because firms will have to follow clear rules without grey areas, which means a boost for free speech”

This is baffling. Social media firms are private: they can (and do) remove whatever they like. The online censorship bill (let’s call it what it is) means the risk lies in publishing, not removing. Her bill will in no way end the arbitrary removal (or shadow-banning) of posts: instead, it will inspire lots more such activity as social media companies privately admit. Rather than have ‘clear rules without grey areas’ the bill has unfathomable rules with grey areas galore, they greyest one of all being the inability to define what is ‘harmful’. Vagueness here is fatal, as it leaves all power to the regulator (who can themselves be overruled by Dorries).

CLAIM
THREE: ‘We’re not turning tech bosses into media moguls. We’re forcing them to keep their promises to users about the behaviour they allow on their sites and giving regulator Ofcom powers to fine them if they fail to do so’.

Her Online Safety bill will absolutely turn tech giants into moguls by asking them – in fact, ordering them – to become censors. To curate which headlines are seen by millions, which they do anyway, but add government censorship preferences into the mix. This is the Chinese model of censorship-by-proxy: Weibo, etc, are given censorship orders and expected to enact. Zuckerberg, Musk etc already have more power than Murdoch, Hearst or Beaverbook ever dreamed of because they will program bots that edit the individual news feed of literally billions of people. The Online Safety Bill gives them power to censor as much as they want without having to say what they have censored, how or why.

CLAIM
FOUR: ‘And finally, yes, the bill rightly has extra protection for journalistic and democratic content – but it’s not just for news outlets. Tech firms have to protect everyone’s free speech and give them a proper right of appeal if their posts are removed.’

Section 16 of the bill seeks to protect journalistic content. But why should journalism have free speech protection which is not afforded to the average citizen? It’s a disgusting proposition. Since when did ministers have the power to decide who gets free speech and who does not? ‘Tech firms have to protect everyone’s free speech’ Ms Dorries says: no, this bill instructs tech firms to censor, not protect. To censor whatever stands a small chance of being seen by her regulator as ‘harmful’. 

Her bill does not define a ‘proper’ right of appeal. Tech firms can still (as they do now) go through the motions. I’ll give an example: when The Spectator recorded Lionel Shriver reading one of her columns it was penalised (demonetised) by YouTube on the grounds that it flouted ‘community guidelines’. (She was criticising lockdown policy.) We appealed: YouTube said our appeal was rejected because our content flouted community guidelines – ie. they didn’t say why. No rationale. No communication. When @PoliticsForAll was deleted by Twitter, no credible reason was given. A private company can publish or remove what it likes. The online censorship bill does not change this, nor define ‘a proper right of appeal’.

SpectatorTV now has 157,000 subscribers and we’re lucky to have good relations with YouTube who I regard as one of the good guys: they didn’t censor claims that Covid was a Wuhan lab leak, for example (unlike Twitter and Facebook). When The Spectator appeals now, we have a direct line to someone we know. I’m grateful for that. But it’s entirely our good luck that we have that relationship. This bill is dangerous because it offers so little protection to the small guys and so much power to the state-run regulator.

As an editor, I’ve had many run-in with tech firms who shadow-ban or censor articles (usually critical of the government). I am actually sympathetic: they’re already expected to scan millions of pieces of content so bot-censorship is the only tool available to them. The humans called in to do this job (Winston Smith would be doing if this 1984 were remade now) are also fallible. Of course BIg Tech will get things wrong: how can it not? But they are responding to the regulatory regime, and in the UK that regime is about to get (a lot) more pro-censorship.

Newspapers are unlikely to defend BIg Tech (who they see as rivals for ad revenue) and government has been sucking up to newspapers by promising to send a ferret up the trouser-leg of Silicon Valley. But in my view, any censorship regime that bans legal free speech will hurt journalism, erode civil liberties and ought to be vigorously resisted. And openly defied if needs be. 

I mean no disrespect to the Culture Secretary when I say that politicians can no more be trusted to regulate the media than foxes can be trusted to regulate a chicken coop. She is creating the most ambitious censorship apparatus in the democratic world – and one that can be used against her party in a heartbeat. She’s the latest of four Culture Secretaries who has been asked to run with this ridiculous Bill (which began life six years ago) and we all know that the law of unintended consequences is the law that Westminster unfailingly passes.

Erdogan and others will be watching open-mouthed. Such censorship is possible in a democracy? You can bully companies, threaten them with fines so they just err on the side of stopping minority voices? And then offer protection to your journalistic friends?

Right now, the Tories are on top. So the ‘accidents’ are happening to far-left outfits (Novara Media, SWP, etc). Such accidents will be made far more common under the new system that she is ushering in – with very little consideration about the long-term consequences.