Daily Mail

Toxic trolls ‘left to thrive on Twitter under Musk’

- By Jim Norton Technology Editor

TWITTER is no longer able to protect users from trolling, disinforma­tion, and child sexual exploitati­on, company insiders have claimed.

Hateful and toxic content is thriving as a result of mass job cuts and changes made under the platform’s billionair­e new owner Elon Musk, they said.

The claims appear to be backed up by academic research showing a spike in accounts following misogynist­ic and abusive profiles and an increase in harassment.

The investigat­ion by BBC Panorama reveals concerns that child sexual exploitati­on is on the rise on Twitter, while there are fewer resources to report suspected offenders to police.

Rape survivors have reportedly been targeted by accounts that have become more active since the takeover, with evidence they have been reinstated or newly created.

One recent employee said targeted harassment campaigns and foreign influence operations that were once removed daily from Twitter are now going ‘undetected’.

At least half of Twitter’s 7,500 workforce has been sacked or resigned since Mr Musk bought the platform last October. Research by the University of Sheffield found a 69 per cent surge in new accounts that were following known misogynist­ic and abusive user profiles.

One insider who works on preventing sexual exploitati­on of children on Twitter said job cuts had undermined their ability to do so – despite Mr Musk labelling it a priority. The employee said his team identified this content ‘daily’ but feared it was no longer all being reported to the police because his staff had been cut from 20 to six or seven.

Twitter said it has removed 400,000 accounts in one month alone to help ‘make Twitter safer’.

The firm’s former head of content design, Lisa Jennings Young, said everyone on her team, which created safety measures designed to protect users from hate, was sacked. Speaking publicly for the first time since, she said one of the most successful features her team designed was the ‘harmful reply nudge’.

Using AI technology to detect harmful language, the ‘nudge’ alerts users before they send a tweet – stopping many people from writing abusive messages.

Another current engineer told the programme that ‘nobody’s taking care’ of this type of work. ‘When you look at it from the outside the facade looks fine, but inside I can see that nothing is working,’ he said.

‘Nothing is working’

Newspapers in English

Newspapers from United Kingdom