The Global Internet Forum to Counter Terrorism, a group of four technology companies – Facebook, Microsoft, Twitter, and YouTube – committed to working together and with governments and civil society to address the problem of online terrorist content, are making a multi-million dollar commitment to support research on terrorist abuse of the internet.
In a blog post on Wednesday, Google’s SVP and general counsel Kent Walker said the forum would look at how governments, tech companies, and civil society can fight back against online radicalization.
The post, a revised and abbreviated version of a speech Kent delivered at the United Nations in New York City, NY, said that the Forum has also set a goal of working with 50 smaller tech companies to help them better tackle terrorist content on their platforms.
“On Monday, we hosted dozens of companies for a workshop with our partners under the UN Counter Terrorism Executive Directorate. There will be a workshop in Brussels in December and another in Indonesia in the coming months. And we are also working to expand the hash-sharing database to smaller companies,” Walker said.
“For our companies, terrorism isn’t just a business concern or a technical challenge. These are deeply personal threats. We are citizens of London, Paris, Jakarta, and New York. And in the wake of each terrorist attack we too frantically check in on our families and co-workers to make sure they are safe. We’ve all had to do this far too often.”
“The products that our companies build lower barriers to innovation and empower billions of people around the world. But we recognize that the internet and other tools have also been abused by terrorists in their efforts to recruit, fundraise, and organize. And we are committed to doing everything in our power to ensure that our platforms aren’t used to distribute terrorist material,” Walker said.
He said that the Forum’s efforts are focused on three areas: leveraging technology, conducting research on patterns of radicalization and misuse of online platforms, and sharing best practices to accelerate a joint effort against dangerous radicalization.
He noted that one early collaborative success is a “hash sharing” database, which allows a company that discovers terrorist content on one of their sites to create a digital fingerprint and share it with the other companies in the coalition, who can then more easily detect and review similar content for removal.
“We have to deal with these problems at tremendous scale. The haystacks are unimaginably large and the needles are both very small and constantly changing,” Walker said.
Walker said that people upload over 400 hours of content to YouTube every minute. “Our software engineers have spent years developing technology that can spot certain telltale cues and markers.
“In recent months we have more than doubled the number of videos we’ve removed for violent extremism and have located these videos twice as fast. And what’s more, 75% of the violent extremism videos we’ve removed in recent months were found using technology before they received a single human flag.”
He added that these efforts are working. Between August 2015 and June 2017, Twitter suspended more than 935,000 accounts for the promotion of terrorism. During the first half of 2017, over 95 percent of the accounts it removed were detected using its in-house technology.
Facebook, he said, is using new advances in artificial intelligence to root out “terrorist clusters” by mapping out the pages, posts, and profiles with terrorist material and then shutting them down.
“Despite this recent progress, machines are simply not at the stage where they can replace human judgment. For example, portions of a terrorist video in a news broadcast might be entirely legitimate, but a computer program will have difficulty distinguishing documentary coverage from incitement. ”