An issue of viral proportions: why are social media companies failing to tackle online hate?

By Alex Jennings

The problems of Twitter are well-known. We hear so often of public figures facing campaigns of vile abuse and threats on the platform, often from anonymous accounts, that such experiences have become normalised. 

Reading recent analysis, however, of harmful content against Harry and Meghan, which revealed a coordinated campaign, where as little as 83 accounts were responsible for about 70 percent of 114,000 often-hateful tweets, it is astonishing how a repulsive minority could manipulate Twitter’s algorithm and continue their activities undetected with such ease.

Emphasising that real humans were behind the campaign rather than automated profiles, the report by Twitter analytics service Bot Sentinel also noted how the accounts deftly crafted material that could fly under Twitter’s detection and suspension system. And with their combined 187,631 followers, Bot Sentinel estimates the accounts were able to reach at least 17 million other users with their tweets.

Twitter is still failing to tackle widespread hate on its platform

Of course, Harry and Meghan wouldn’t have received this latest hate, since they quit social media almost a year ago precisely because of such abuse. There is no clear goal of the slander, either, unlike with the campaign to have actor Amber Heard dropped from the Aquaman sequel, following her allegations of abuse against ex-husband Jonny Depp. However, whatever the motive for this latest campaign, it undoubtedly highlights not only how Twitter is still failing to tackle widespread hate on its platform, but also how challenging that mission is proving to be.

Following a year in which England footballers received a torrent of racist abuse and public figures, especially MPs, have received ever more threats, it is clear that the status quo is unacceptable. But while regulation of social media companies is long overdue, there are certainly no quick fixes.

The government’s Online Safety Bill, which is in its final drafting stages, seeks to give the communications regulator Ofcom power to fine companies up to 10% of their annual global turnover if they fail to remove harmful but legal content from their platforms. The proposed legislation may be a step in the right direction towards ending self-regulation by media companies, but it is far from perfect: ‘harmful’ is a subjective term, opening the door to politicised regulation, as government ministers could influence what Ofcom should and shouldn’t define as an infraction. How is Ofcom supposed to monitor the volume of content on a platform such as Twitter, with its 200 million users?

Proof of identity could nevertheless be a mostly positive compromise in making social media a safer space

Indeed, Twitter itself relies on automated algorithms to sift through millions of tweets and detect offensive material, since the 1,500 human moderators it employs would otherwise be unable to keep up with demand. But if the actual aim of the government’s regulation is to force social media companies to take online abuse more seriously and use better technology to stamp out hate, they may be disappointed. 

True, a company such as Twitter, earning over $2.7 billion in revenue last year, has huge resources at its disposal and could always do more. But automated technology, while apparently successful in censoring illegal images and incendiary hashtags, is still a blunt instrument when it comes to seeking out abusive messages. Algorithms are not – and may never be – sophisticated enough to understand context, nuance and cultural specificities; while a blanket ban on many potentially controversial words or phrases, like on Chinese social media, would be inimical to free speech.

An end to online anonymity, though, which many MPs have called to be added to the proposed legislation, would certainly prevent many people from being emboldened by online platforms to offend without fearing accountability. Although operating under a pseudonym can be legitimate and indispensable for many, including victims of domestic violence and whistle-blowers, proof of identity could nevertheless be a mostly positive compromise in making social media a safer space.

The emotional toll we hear of from victims cannot be overstated

Some commentators have argued that ultimately it is the persistence of racism and hatred in society that is the real problem here, with social media merely exposing the existing issue. They are not wrong, but such poison cannot be allowed to proliferate through social media. 

Something must be done too about the harassment and abuse which public figures and many others face online: the emotional toll we hear of from victims cannot be overstated. That’s why government regulation, despite not being a cure or even a revolutionary treatment, could at least be a starting point in fighting the dreadful disease of online hate.

Image: Brett Jordan via Unsplash