Delivered to the George Washington University Centre for Cyber and Homeland Security's extremism programme.
Good afternoon. I would like to thank Dr Vidino, Dean Knapp and the team here at George Washington University for convening this important gathering to discuss building a global partnership to combat extremism online in all its forms.
It is indeed ironic that one of humanity’s most liberating innovations, the internet, is being misused in this way - as a vessel for violence and hatred.
Over the past few years, I have spoken at many symposiums about the escalating threat posed by the abuse of internet platforms and applications by terrorists and extremist groups. I have called for united action and spoken about the urgent need for governments, civil society groups and the private sector to come together to apply new technologies, share information and develop best practice solutions. But today we are at an impasse.
It is no longer a matter of speculation that terrorists and extremists use internet platforms and applications to inspire violence, spread extremist ideology and to plan and execute attacks. Each tragic incident reconfirms it.
We come together this week following yet another horrific attack. This one in London around the Houses of Parliament; my home and my place of work. Five people tragically lost their lives, 50 were injured and 30 hospitalised. The victims were from 12 different countries. And within 24 hours, Daesh claimed responsibility.
We know how extremists manipulate information and sow the seeds of discord in society. We know how they use propaganda to reinforce grievance and instigate hatred. And we know how they convince people to give up their lives and join ‘the fight’. They target those who feel vulnerable, marginalised and invisible. But today the pool is expanding exponentially, as billions connect to social networks - fertile platforms for enticing and enlisting recruits.
Yet unlike the physical world where national governments can take clear and firm actions to keep people safe and secure in their homes and communities, the virtual space is the domain of commercial companies and we must rely on their cooperation and support to keep people from harm.
If we are to protect human life in this ever more connected world, we need a new model of shared responsibility and this is the conversation I want to have with you today. I would like to challenge our thinking and ask how we move from reacting to crisis, to prevention and a full acceptance of responsibility on all sides.
In western societies we hold sacred our democratic values: freedom of speech, the right to privacy, the rule of law, safety and security. These values apply to the internet as well, which we believe must be free, open and accessible by all. But there is growing public sentiment that not enough is being done to tackle terrorist misuse of the internet. And when it comes to the question of what to do about it, we are at an impasse.
But the voices of consumers and brands are loud and clear and the recent exodus of top advertisers from social media platforms sends the strongest possible message that their products and services must not be promoted next to deplorable extremist content.
Recently, Germany proposed legislation that sets out binding standards for how social network companies should delete criminal content. However, the challenge with this approach is that regulation creates a new set of national rules for these businesses that are by definition global and borderless. And there is significant complexity around any regime that governs online activity not least keeping any such obligation current given the speed and evolution of technology and extraterritorial jurisdiction that applies.
So, we are at a critical moment, when united action to tackle this threat is the only way forward. Governments and experts can provide extensive knowledge and a rigorous understanding of the threat but industry is best placed to innovate on technical solutions that address this threat specifically for their own commercial platforms. They must innovate and automate their response to identifying and removing this vile, hateful material so that together we can ensure that everything possible is done to stop it infiltrating and poisoning a global audience.
And increasingly we see a potent ‘cycle of hate’ across multiple groups and ideologies as all sides of the extremist spectrum feed off each other, escalating tensions. We saw this post the London attack when this image appeared. The photographer who took this picture said he posted it because it showed the young woman wearing a hijab was traumatised by the events around her. But this was not how she was portrayed. The image was wildly mis-appropriated by right wing groups as representing her as insensitive and indifferent to the carnage that was unfolding.
The growing audacity of all types of extremist groups is perpetuating this vicious cycle. The far-right asserts that Islam represents an existential threat to the West, stating that all Muslims are supportive of terrorism. We have seen terms like ‘rapefugees’ used in social media, stigmatising those fleeing the atrocities in the Middle East. The far-right is using groups like Daesh as an opportunity to frighten, sow division in communities, and make their extremist narratives more palatable.
Social networks algorithmically connect like-minded individuals and amplify their passions. That is the core of the online advertising business model. But these connections can channel people into echo chambers where highly emotive and passionate content, amplified by these algorithms, reinforces extremist messaging. This creates an illusion of strength in numbers when these views are in fact fringe views.
Over the past year, we have seen new developments such as the use of social media live during attacks that complicate disruption efforts. Terrorists document their unspeakable actions and bask in a nihilistic personal moment of fame and notoriety. Furthermore the footage is later released and used in instructional videos to inspire and incite more violence, perpetuating a vicious cycle.
The precedent for the real-time sharing was tragically set in 2016 in the broadcast of the murder of a French policeman and his wife on social media while their little boy looked on in horror. We saw this used again a few weeks ago during the horrific attack in Kabul on a military hospital, where the attackers posted live photos as the event was unfolding.
Terrorists’ use of the internet as a sphere of influence will continue to evolve and adapt, and we need new methods to quickly identify and remove terrorist and violent content, and to deliver more effective strategic communications to counter these deadly narratives. And we must be evidence based; too often we are reactionary and do not adequately deal with the complexities faced.
Daesh in particular produces material designed to capitalise on community tensions and hostility towards Muslims in the West. This further isolates groups away from mainstream society. Their message is that every Muslim has a duty to fight - that ‘jihad’ is a fight that is local, as well as global and that if they are unable to travel, there are legitimate targets in their home countries.
Following the attack on Berlin’s Christmas market, Daesh released a statement on Telegram in Arabic, French and English encouraging their supporters to carry out lone actor attacks in the West and Europe during the holiday period, specifically advising supporters to target ‘celebrations, clubs, hospitals, markets and movie theatres’. Tragically, foretelling the Reina nightclub attack in Istanbul on New Year’s Eve.
We have seen this call to arms before. Last year, the then spokesperson Abu Muhammad al-Adnani, called for the group’s supporters to carry out terrorist attacks during Ramadan. At the time, analysts regarded this call as less persuasive than his previous fatwas. But in fact it proved deadly. It was the bloodiest Ramadan this century.
Other emergent narrative themes in Daesh propaganda include the total rejection of LGBTQ communities and a hatred of nationalism and secularism were cited as motivation for the attack on the Pulse nightclub in Orlando and the lorry attack in Nice on Bastille Day.
This directional shift is reflected in Daesh’s official online publications as well. The image on the left of the screen shows its Dabiq magazine which encouraged supporters to embrace the Caliphate and build a new state, the second shows its newer publication, Rumiyah which calls on its followers to carry out acts of terrorism wherever they live and wherever they can, reflecting the new reality and its survival as an ideology.
Let me illustrate some of the atrocities that Rumiyah has inspired, as each edition contains graphic instructional videos. The second edition called for lone actor attacks using knives with a full demonstration on a frightened hostage, which was replicated in the heinous murder of the French priest in Normandy and attacks in the Minnesota shopping mall last year. The third edition, advised it supporters to carry out vehicle-based lone actor attacks, praising the Bastille Day Nice attack replicated in Berlin and here in the US at Ohio State University.
Recently, a scene showing how to build a shrapnel-filled IED in a kitchen was disseminated. More than 100 links to this video were posted across 29 platforms within an hour. This was organised on Telegram, distributed on Twitter and the video was hosted on YouTube, Archive, Send Vid and Google Drive, part of the terrorist ‘ecosystem’ used by Daesh to ensure their propaganda has maximum impact. Twenty-four hours after the video’s release, despite best efforts, half of these links were still active.
Although their ability to produce and distribute propaganda has declined, their ability to reach a global audience has not. According to research the UK government has undertaken, as Daesh have been degraded and defeated, and their infrastructure compromised on the ground, its ‘unofficial’ brand ambassadors are prominent distributors of propaganda. These devotees critically create their own material which they promote and discuss across networks. Idolisation of lone-actor terrorists as martyrs for the cause feeds an increasing audacity in attacks and offers members the chance to gain fame, status and glory.
We now see other terrorist groups mimicking Daesh online tactics with devastating consequences. In fact we see an escalating competitive dynamic amongst terrorist groups – a savage game of one-upmanship.
In the past few weeks we have seen the Syrian al-Qa’ida linked franchise Hayat Tahrir al-Sham (HTS) launch al-Ebaa, a professional media brand for its online communiques and videos. In mid-March they released their first English language statement, rebuking the US for its ‘selfish’ policy on Syria.
As the quality and quantity of Daesh output fluctuates, we must not lose sight of other extremist groups seeking to increase their online presence. In the UK, the neo-Nazi group National Action were proscribed for glorifying terrorism, having built up a fan base through the use of explicitly youth-orientated material to reach new recruits. The ability for other terrorist groups’ propaganda to incite violent attacks is no longer theoretical as the brutal murder of Jo Cox, a UK Member of Parliament, last summer demonstrates.
The threat we face continues grow. We must develop and rapidly deliver an ever-stronger response at pace and scale. Some progress has already been made.
In the UK, we have developed a world leading approach for tackling terrorist and violent extremist use of the internet focuses on 2 areas of work:
Working with industry to voluntarily remove extremist content online through the counter terrorism internet referral unit.
Bringing communication experts and civil society groups together to develop and run targeted and effective counter messaging campaigns that provide compelling alternative voices to extremist rhetoric.
Working together with industry in 2016, the UK’s counter terrorism internet referral unit, run by our metropolitan police department, secured the removal of over 120,000 pieces of terrorist and extremist content. We supported the setting up of this model in the EU, and their unit was launched in July 2015 to secure the removal of content in a wider range of languages. The unit has reported that 90% of their referrals to industry have been removed.
Following the Paris attacks, Telegram acted swiftly to suspend the accounts of 78 public channels used by Daesh and supporters in 12 languages. Telegram recently to the UK request to remove English language Daesh propaganda. And last year, we saw leading internet companies come together in December in a proposal to build a shared hash database of terrorist content at the EU Internet Forum.
So this looks like progress, right? Yes, but it is still too little too late. By the time we react, the terrorists have already reached their audience. Research conducted by the UK government shows that the majority of links to terrorist content are shared within 2 hours of first release. They anticipate take-downs and suspensions by instructing their supporters to return to the open net time and time again.
This must change. We need a new approach, a new partnership. Governments across the world are agitating, looking at legislation to force the more timely removal of content by social media companies; reinforced by fines and other sanctions. I do not believe this approach alone will succeed.
It is incumbent upon industry to drive this change. They must build new capacity that is holistic, targeted and dynamic to address these threats and reclaim their platforms from those who exploit them, incite violence and push dogma and repression.
Finally, if the terms and conditions that govern these sites are based on corporate philosophies, values and beliefs, then surely the goal posts can be moved.
While violent extremist groups seek to undermine the very ideals and values that the internet was established to advance, we must reinforce its capacity to be the answer to hatred and intolerance, rather than the vehicle for it.
We must understand the influence of terror groups online and deploy the use of strong data analytics to recognise how terrorists amplify their messages and to decipher the patterns of their deadly actions.
And in terms of technology, we need to improve solutions that classify the language of extremism, automate the identification and removal of dangerous content at scale, and create tools that better tackle automated bots and other techniques that support these propaganda machines. In essence we must take a stance of solidarity against terrorism. Solidarity across governments, industries and peoples around our world.
The internet is universal and it is imperative that developers consider these dangers alongside other internet harms. We need companies to champion and support projects that build digital resilience - programmes that help young people think critically about what they see and read online so they can make informed and safe choices. Together with online safety charities we can help increase awareness, confidence and capabilities.
We must empower the global community with better tools to report and respond to harmful content; to speak out and take action. Every person has the ability to recognise bias, hatred, and intolerance and to say, no, not on my profile, not in my name.
We can’t let the world retreat to a dark place of ignorance and prejudice. We must stand up for what we believe in. Freedom. Peace. Democracy. Understanding. Inclusivity. A world in which knowledge, debate and discussion bring people closer together and make them feel a part of something greater than themselves.
To close, I will leave you with this quote from Prime Minister Theresa May the day following the London attack:
Yesterday we saw the worst of humanity, but we will remember the best.