YouTube Turns to Improved Machine Learning in Fight Against Terror Videos

Google's YouTube said Tuesday that it's increasing efforts to fight terror-related content on its site. The move comes roughly two months after British Prime Minister Theresa May called tech companies hotbeds of extremist recruiting following an attack on London Bridge that killed eight and injured 48. It also follows a springtime marketer revolt over ads' unwanted proximity to hateful and violent video on YouTube. YouTube now claims it will use improved machine learning technology to better detect and remove terror-related videos from its video streaming platform. It's also brought on more than a dozen different expert groups, will enforce tougher standards and has increased its role in creating counter-radicalization content, it said in an update about its efforts. The video streaming giant said its machine learning tech is faster and "more effective than ever before," adding that 75% of videos that it removed for violent extremism over the past month were taken down before humans could flag them. "With over 400 hours of content uploaded to YouTube every minute, finding and taking action on violent extremist content poses a significant challenge," YouTube said on a blog. "But over the past month, our initial use of machine learning has more than doubled both the number of videos we've removed for violent extremism, as well as the rate at which we've taken this kind of content down." The groups tapped by YouTube include the Anti-Defamation League, No Hate Speech Movement and the Institute for Strategic Dialogue. They will bring "expert knowledge" around issues like hate speech, radicalization and terrorism, according to YouTube, that will help it identify questionable content. Jigsaw, an incubator within Alphabet that builds technology to tackle various security challenges facing the world, including ISIS-fueled extremism, has started rolling out features using the company's "Redirect Method" to YouTube. People who search for sensitive keywords on YouTube, for example, will instead see a curated list of videos that attempt to debunk and directly confront violent messages. YouTube also hopes it can persuade would-be terrorists from joining groups like ISIS by creating workshops that create a "positive sense of belonging online." Despite the negative press around big marketers' ads running on terror video, Google's bottom line appears intact. Then again, Procter & Gamble said in its latest results that it had cut its digital advertising by $140 million in the most recent quarter over "brand safety" concerns and still outperformed analysts' forecasts and its rivals. So the long-term effects on YouTube are far from settled.

YouTube Turns to Improved Machine Learning in Fight Against Terror Videos

Featured Stories