Twitter is coming under increased pressure to eliminate explicit child sex abuse images uploaded on the popular social media service as child protection advocates complain the process is taking too long.
Dozens of child sex abuse images are uploaded and shared on Twitter and in some cases have remained online for as long as a year. This is despite user reporting systems and the recent addition of tagging software.
Australian media personality Charlotte Dawson used her large Twitter following last week to suspend a public account posting images of young children being raped. But Ms Dawson claimed it was hours and hundreds of tweets before the account was removed.
An independent Fairfax Media investigation identified dozens more images of children, some who appeared to be as young as five, being sexually abused. Fairfax has since reported these accounts to Twitter.
Many of the accounts had been suspended even as they were being reported, but others remained.
The images of abuse were shared as public posts between groups of adult user accounts, attracting commentary as they were shared. One user, whose profile described him as ”loves little boys come play with me” sent a public message to Twitter last month saying: ”@twitter hi why can’t you leave my friends alone if they want to post pics of kids ppl that is their business knot urs don’t do that 2 them.”
Twitter user Christabelle Oblowski, who is based in London, said she was able to identify nearly 40 active accounts sharing child exploitation images within minutes. ”All with very obvious pictures of young girls [preteen] in various poses and states of undress, up to and including full nudity and sexual contact,” she said.
A user she followed had launched a public attack on a man who claimed to be posting pictures of his naked 13-year-old sister. ”The pictures were quite obviously of an under-age girl, and they were uploaded in August 2012,” she said. ”The account had almost 1000 followers.”
She reported the URLs to Twitter and half were suspended within 24 hours, some were taken down a few days later, but others were left up.
Twitter announced plans to adopt Microsoft-developed PhotoDNA software, which identifies images of abuse through a tagging system, in July. Facebook has been using the software since 2011 and Google has been using ”hashing” technology to tag known child sexual abuse images since 2008. A Twitter spokeswoman said the PhotoDNA software had ”recently” gone operational on its service.
”Clearly it’s failing then,” said CyberSafetySolutions’ Susan McLean. ”PhotoDNA is meant to identify that stuff behind the scenes and it’s not doing that.”
The Twitter spokeswoman said the company was working with Microsoft and Thorn to find and remove images of child sexual exploitation.
”When we are made aware of links to images promoting child sexual exploitation, they are removed from Twitter without further notice and reported to the National Centre for Missing & Exploited Children,” she said.
Ms McLean said Twitter had made huge improvements over the past year, but Ms Dawson’s experience highlighted lingering deficiencies.
Users who find child exploitation material on Twitter are urged to email firstname.lastname@example.org and include links to the accounts and relevant tweets.