Facebook, Google and Microsoft team up to build AI paedo-hunter that tracks down child abuse online
TECH titans including Facebook, Google, Twitter and Microsoft are teaming up to thwart paedos online.
The group has announced Project Protect, in which the companies will work together to stop child sex abuse content appearing on the web.
Speaking as part of the existing Technology Coalition of digital firms, the group outlined a "five pillar" plan for action.
Firms will work together to push innovation in technology to detect and stop child sex abuse content, encourage more collective action, fund more independent research, increase information sharing and increase transparency and accountability.
Last week, industry experts told MPs on the Home Affairs Select Committee that social media firms were still not doing enough to stop the spread of online child abuse.
Robert Jones, from the National Crime Agency (NCA), and Susie Hargreaves, from the Internet Watch Foundation (IWF), called on the sector to do more and argued there was no excuse why more was not already being done given the detection technology available to the major platforms.
Mr Jones told MPs the response of social media firms was being "hampered by a lack of regulation".
Both he and Ms Hargreaves said the Government's proposed Online Harms regulation, which would have greater penalties for firms who fail to protect users, was vital to improving this response.
Project Protect will establish a multi-million pound research and innovation fund to support the building of new tools to prevent the spread of child sex abuse content.
It will also create a new forum of experts and a commitment to publish annual progress reports on industry efforts.
The Technology Coalition, formed in 2006, is a partnership of 18 technology companies of various sizes, which includes tech giants Amazon, Apple, Facebook, Google, Microsoft, Twitter and Snapchat.
Home Secretary Priti Patel said she welcomed the steps taken by the tech firms involved, after several signed up to a collection of voluntary principles around online safety earlier this year.
"The sexual abuse of children online is sickening and we must all work collaboratively to eradicate this crime.
"I welcome this step from the Tech Coalition and hope this drives meaningful change to protect our children," she said.
"The voluntary principles to counter online child sexual exploitation and abuse is a landmark blueprint to keep our children safe online.
"Myself and the Five Country partners have been clear that technology companies need to work quickly and go further to address the critical issues that could leave children vulnerable to online predators."
In a blog post announcing the plans, the group said advances in technology and social media had "added to the challenge of keeping the internet a safe place" but after consultation with experts had "renewed investment and ongoing commitment to our work seeking to prevent and eradicate online CSEA".
The coalition said it was also working with The Global Partnership to End Violence Against Children (EVAC) and the WeProtect Global Alliance as part of the project.
On the new plans, IWF chief executive Ms Hargreaves said: "I'm delighted to see a renewed commitment by the technology industry to fight child sexual exploitation and abuse online with the launch of Project Protect.
"It's imperative that companies come together and fight this in partnership in order to gain real results, for children."
Google senior vice president of global affairs Kent Walker said the scheme will help firms share "progress, learnings, and cutting-edge tools" to help fight the problem, adding that "no company can fight this problem alone".
Sheryl Sandberg, Facebook's chief operating officer, said the project "brings together the brightest minds from across the tech industry to tackle a grave issue" while Twitter's head of trust and safety Del Harvey said the firm welcomed the "renewed effort to collaborate with our peers".
Facebook is already using computer systems to spot suspected paedos – and warn underage users who may be being preyed upon.
The new system is being rolled out as part of an update to Facebook's hugely popular Messenger chat app.
Facebook is using an AI machine learning system to spot potential abusers in advance.
Who is on Facebook's Safety Advisory Board?
Here's the official list...
- Centre for Social Research (CSR)
- Childnet International
- Connect Safely
- The Family Online Safety Institute (FOSI)
- Insight SA
- iWIN
- National Network to End Domestic Violence (NNEDV)
- Net Family News
- Netsafe
- PROJECT ROCKIT
- SaferNet Brazil
The system will scan for suspicious behavioural signals.
This includes an adult sending a large number of friend or message requests to people aged under 18.
When this happens, users under the age of 18 will be alerted, with a warning urging caution.
Safety notices will pop up in a chat and provide tips to help users spot suspicious activity.
And users will be warned to avoid interacting with adults, and to block or ignore dodgy users.
If these machine learning systems detect very sinister behaviour, Facebook moderators can go on to delete the account.
Most read in Tech
In other news, find out how to make Facebook Avatars that you can model after yourself.
Facebook and Instagram have launched virtual shops where you can buy almost anything.
READ MORE SUN STORIES
A recent surge in users means that 2.6billion people now log onto Facebookevery single month.
And a controversial new Facebook design is rolling out to everyone.
We pay for your stories! Do you have a story for The Sun Online Tech & Science team? Email us at [email protected]