From social networks to speech regulators

Defending their client from a defamation claim in Texas last week, the lawyers of Alex Jones, a prolific American conspiracy artist, made a surprising assertion. Despite his lucrative media empire, billed as “seeking the truth” and “exposing lies” (as well as selling dietary supplements), they argued that “no reasonable person” would interpret one of his claims as “a statement of fact”. In other words, don’t take him seriously.

But millions of people do take Alex Jones seriously – including a target of his who was forced to move house seven times, and her stalker, ordered by a judge not to watch any more of his shows after repeatedly threatening her life. Now, finally, the big web platforms have woken up too. On Monday, after weeks of prevarication, Facebook, Youtube and Apple all banned him for violating their codes on hate speech.

Better late than never, perhaps. But the timing shines a spotlight on what a mess tech companies are making of policing their own platforms. Mr Jones has been operating since before 9/11 (an inside job, he says), so why has he only just been banned? And if it was due to a fair and consistent policy, rather than to save face, why did all the companies act at once – with Facebook and Youtube rapidly following Apple?

Traditionally, social networks have defended themselves from demands to remove content by talking about free speech. When Facebook’s John Hegeman was asked recently about Mr Jones, he said: “Different publishers have very different points of view.” Mark Zuckerberg said he won’t censor Holocaust deniers because they’re not “intentionally getting it wrong”. Twitter, which has not banned Mr Jones, said in 2012 that it was “the free speech wing of the free speech party”. One problem with this argument is that it’s not true. None of these companies is shy about enforcing their own rules when they see fit. Youtube uses AI to purge copyrighted material. Facebook once mercilessly excised pictures of female nipples, regardless of their context. Twitter regularly suspends users who tweet swear words at famous people. That suggests a deeper problem: that for social networks, free speech is a tactic and not a principle. Claiming to uphold it allows you to avoid all sorts of difficult questions and decisions. It allows you to leave as much content online as possible, thereby selling as many ads as possible. It allows you to keep intact your business model, which is based on outsourcing most tasks to your users and automating nearly everything else, without having to hire masses of lawyers and moderators.

I sympathise with this reluctance. To begin taking hate speech or fake news seriously is to enter a terrible ratchet. Every time you intervene in a political debate, you will help one side and hurt the other. You will then face pressure to intervene in the opposite direction in future. Soon both sides will regularly demand that you intervene for them, and every time you do you will strengthen the expectation that you will do so in future. No social network wants to be in that position, and frankly none of us should want that either.

Unfortunately, it’s already too late. Tech firms have already assumed the power to control what we see and say in pursuit of advertising dominance. On a basic level they incentivise certain kinds of speech over others, warping the way we communicate. It is their own algorithms that spread fake news because it provokes strong emotions; their own recommendation engines that relentlessly guide new audiences towards fantasists like Mr Jones by taking the devotion of his faithful as an endorsement.

Belatedly, they are trying to fix this, but in the meantime Western publics have quite reasonably seen all this and demanded tech firms do something. None has been able to resist that pressure, but none has undertaken root and branch reform either. The result has been flailing inconsistency: complex guidelines enforced so arbitrarily as to destroy users’ confidence, with exceptions carved out for popular users who are only banned when scandal becomes unmanageable.

There are, logically, three options ahead. One is to keep careening between censorship and laxity. Another is to genuinely become free speech fundamentalists – resist any demands to censor short of legal force, and weather the massive criticism that would result. The last is for tech firms to take seriously the position they have voluntarily seized as de facto speech regulators, and to enforce their codes consistently and transparently. That would require a massive, permanent increase in the resources they devote to moderation. But since they will probably be forced to do this anyway, they might wish to cut to the chase.

‘Every time you interfere in a political debate, you will help one side and hurt the other’