IAN HERBERT: If big tech giants do not use AI to unmask trolls then they are as good as condoning it

We’ve spent the last week hearing the social media giants’ assurances about all the work they’re doing to protect players from abuse. 

Twitter says the number of accounts it has taken action on in the last years has increased by 105 per cent. Without actually revealing what that figure might be.

So I undertook a random 30-minute check, today, of accounts which have issued racist, homophobic abuse directed at players. 

Twitter and Facebook have told Premier League clubs they will never end the practice of allowing users to open anonymous accounts

Twitter and Facebook have told Premier League clubs they will never end the practice of allowing users to open anonymous accounts

Players such as Arsenal duo Granit Xhaka (left) and Hector Bellerin (right) have been targeted by social media trolls

Players such as Arsenal duo Granit Xhaka (left) and Hector Bellerin (right) have been targeted by social media trolls

Only to discover that the perpetrators are untouchable. Despite the hate they have disseminated, they and their accounts were, as of Tuesday, still up there in the Twittersphere. Larger than life.

There’s @Lord_Reaper_ for example, who also goes by the ‘name’ of ‘Your Worst Nightmare’ on Twitter and can list a particularly vicious piece of homophobic and racist abuse for Arsenal’s Hector Bellerin among his creative output. (I won’t give these reptiles the satisfaction of relating their abuse.) 

He’s still there, up there in plain sight, on Twitter. He was tweeting only last week.

There’s Nicholas Griffin, who has racially abused Davinson Sanchez. Nicolas seems to be a teenager, wearing a Manchester United shirt and taking a picture of himself for his Twitter profile. He’s still up there too.

There’s Niall Burke from Bolton, pictured with a new-born for his own initial Twitter profile, though the baby’s image been removed now. He’s being careful. His Twitter account is protected, so only ‘approved followers’ can see his tweets. Just as well, considering the foul homophobic abuse he unleashed on Bellerin.

And there’s Ronny Wheeler, who said he wanted harm to be inflicted on Arsenal’s Granit Xhaka, after the player was sent off during Arsenal’s defeat to Burnley. Ronny’s still there, too. Twitter profile protected. Doesn’t want to be getting abuse.

Bellerin and Xhaka were singled out for homophobic, racist and violent abuse and threats

Bellerin and Xhaka were singled out for homophobic, racist and violent abuse and threats

There are others, like @_cxllumD_ and @caesar_TheF. Not the kind of people you ever would want your children or your parents to meet.

The online existence of such cretins disgraces the tech giants, who say they are horrified by the abuse and to have no room for it. Who throw out big numbers about the ‘content moderation’ systems which are supposed to detect hate and ban the perpetrators.

The names tumble out from the work of data science company, Signify Group – which worked with the PFA on a major study which last year revealed the scale of abuse.

Using an artificial intelligence software package it calls Threat Matrix, the firm has shown it is perfectly possible to analyse hundreds of thousands of messages and pull out abuse aimed at footballers – be-it texts or emojis. 

This graph shows the level of abuse directed at Arsenal players ahead of their game at Spurs

This graph shows the level of abuse directed at Arsenal players ahead of their game at Spurs

One of the homophobic messages directed at Bellerin after the game with Spurs

One of the homophobic messages directed at Bellerin after the game with Spurs

Six posts actually called for harm to Granit Xhaka after his red card during the Burnley game

Six posts actually called for harm to Granit Xhaka after his red card during the Burnley game

The firm’s findings are depressing. They claim to have identified 3,500 abusive messages. The likes of Nicholas, Ronny, Niall and Lord Reaper are a few of many.

The firm claims that in up to two-thirds of cases, data can be cross-referenced with public information and provide more detail on the abusers – sometimes even establishing whether such an individual is a season-ticket holder at a club. 

No fewer than 25 abusive comments, some of them racist, have been detected from football club season-ticket holders, Signify says.

The abuse emerges from a Signify analysis of tweets leading up to specific Arsenal and Chelsea games and sent in a 30-day period to Antonio Rudiger’s account. The findings were revealed by Mail Online’s Charlie Walker on Tuesday morning. Twitter have not yet responded.

Signify analysed the messages sent to individual Spurs' players around several matches

Signify analysed the messages sent to individual Spurs’ players around several matches

The leaders of English football sent a letter to Twitter chief executive Jack Dorsey and Facebook CEO Mark Zuckerberg (pictured) about levels of abuse on their platforms

Twitter (pictured is CEO Jack Dorsey) will, like Facebook, continue to toughen up the policing of social media abuse in-house

The leaders of English football sent a letter to Twitter chief executive Jack Dorsey (right) and Facebook CEO Mark Zuckerberg (left) about levels of abuse on their platforms 

The tech giants all say that they’ve got this problem handled. They consider themselves to be guardians of freedom of speech. 

That’s why they’re comfortable to have rejected, out-of-hand, the English football authorities’ request last week that abuse be blocked before it is sent. ‘Pre-moderation’, as it’s called.

It’s the users of the platforms, not the operators, who are the racists, of course. But the fact that abusers are out there, utterly uninhibited and easily detectable, shames those who run the networks. 

The platforms are as good as condoning this abuse if they do not use the technology to eradicate it.   

source: dailymail.co.uk