Inside the dystopian ‘HateLab’ that’s surveilling social media

Tackling so-called hate speech has become an obsession of the Western elites. A whole host of criminal laws, regulatory bodies and NGOs have sprung up in recent years, all of which are working to suppress the expression of views that some deem offensive.

One such initiative is Cardiff University’s ‘HateLab’. Established in 2017 as a joint venture between criminologist Matthew Williams and computer scientist Pete Burnap, the HateLab is committed to ‘identifying’ and ‘monitoring’ any and all hateful speech online. It claims that, through the use of statistical modelling and machine learning, it can track trends in online ‘hate’ in real time. The lab in Cardiff even has a dashboard that displays these trends for the UK. The Observer describes it as an ‘all-seeing eye for the racist, misogynistic, homophobic, transphobic venom that humans spit at one another from behind keyboards’.

For Williams, one effective way to challenge ‘hate’ is through humour. ‘Engaging in a lighthearted way and maybe being a bit sarcastic’ with the ‘haters’, he suggests, can ‘highlight the inconsistency of their argument’.

If this were the limit of HateLab’s activities and influence, then there would be little to object to. Mocking and ridiculing bigoted views is in the fine tradition of counter-speech – that is, the idea that the best remedy for bad speech is more speech and better arguments.

But the HateLab does more than just draw attention to the hateful views that we might want to rebut. The real-time data it generates is also being used to ban and restrict people’s speech. For instance, research conducted by the HateLab has been ‘embedded’ into the UK’s National Online Hate Crime Hub, which empowers the police to ‘address’ (ie, criminalise) online hate speech. Its research has also helped inform the Welsh government’s approach to hate crimes.

The HateLab’s data has even been used beyond the UK, by various NGOs, Big Tech companies and other national governments. The US Department of Justice has provided a grant of around $900,000 for the HateLab to research a link between hate speech online and hate crimes in the real world. Google, one of the largest tech giants, helped to fund the machine-learning research behind the HateLab’s real-time monitoring capabilities. These are major investments, from governments and giant corporations, in technology that can be used to monitor and curtail online speech.

This mission to stamp out online hate might sound benign. But hate speech is an inherently subjective category, and so empowering the authorities to censor it is extremely dangerous. Indeed, the scope of what can be deemed ‘hateful’ now goes well beyond obviously racist or bigoted views to encompass entirely mainstream political opinions – from support for Brexit and Donald Trump to a belief in biological sex.

HateLab founder Matthew Williams goes as far as to claim that the 2016 votes for Brexit and Trump were ‘key drivers of racist attacks’. Of course, as has previously been shown on spiked, the supposed surge in Brexit-driven hate crimes is a myth. It is a product primarily of changes to the way hate crimes are recorded and of a concerted effort from the authorities to encourage the reporting of ‘hate’. The media’s fear-mongering over this non-existent wave of hate was a flagrant attempt to delegitimise the Leave vote – to associate a demand for democracy and sovereignty with racial abuse.

As well as this hate-crime panic, the populist revolts of 2016 also provoked an elite panic over online ‘misinformation’. As the Twitter Files revealed last year, Big Tech, governments, NGOs and academia colluded to root out and censor certain views on the (often spurious) grounds that they constituted misinformation. Again, the partisan bent of this was clear. Information that could be seen as benefiting Trump’s 2020 re-election campaign was far more likely to be hidden from view – even when it was true, such as the New York Post’s scoop about Hunter Biden’s laptop. Similarly, during the Covid pandemic, opposition to lockdowns, masking and vaccine mandates was curtailed on social media for breaching misinformation rules. Even prominent scientists and reports of peer-reviewed scientific trials were censored for challenging government Covid policy.

What all this shows is that it is not just racists or bigots who have something to fear from the real-time monitoring and surveillance of ‘hate speech’. The HateLab’s work fits into a wider elite agenda to deligitimise dissenting views as ‘hateful’ and to have them scrubbed from the internet. The danger this poses to free speech must not be underestimated.

Thomas Osborne is an intern at spiked.