Counter-terrorism was never meant to be Silicon Valley’s job. Is that why it’s failing?

Counter-terrorism is being slowly privatized and carried out by low-paid workers at technology companies such as Facebook, Google and Twitter.

Although these companies hire expert advisers and former government agents to tackle extremist propaganda and recruitment enabled by their platforms, much of the grunt work is carried out by contractors earning $15 an hour or, in YouTube’s case, volunteers.

The result is a private counter-terror workforce with little training increasingly employed to do the kind of work expected of law enforcement. Such work is carried out secretly, inaccurately (journalists and activists have been censored) and with little accountability.

On Monday, Facebook, YouTube, Twitter and Microsoft announced the creation of a joint forum to counter terrorism following years of criticisms that the technology corporations have failed to block violent extremists on their platforms.

However, the current existence of secretive teams set up to combat terror on social media platforms already highlights an awkward bind for the technology companies. On one hand they want to be seen to protect their users’ privacy from government intrusion. But at the same time they are under great pressure to prevent their platforms from being used to radicalize and incite violence.

Leaders including British prime minister Theresa May and French president Emmanuel Macron have labeled these platforms as breeding grounds for terrorism and have demanded they build better tools to identify and remove extremist content. The European Union has created a draft law to make social media companies crack down on extremism or face steep fines.



Isis propaganda video calling on US-based radicals to carry out terror attacks. Photograph: YouTube/Eric Feinberg

“It has become the responsibility of the tech giants, against their own will, because law enforcement are simply not able to do this,” said Carl Miller, Research Director of the Center for the Analysis of Social Media at think tank Demos. “This is partly a law enforcement capacity issue and partly because tech giants don’t want to give states access to large amounts of data.”

Twitter, Facebook and Google have announced measures to tackle the problem, including suspending accounts, using artificial intelligence to identify extreme content, hiring more content moderators, developing and supporting “counter-speech” campaigns, and creating a shared industry database of hashes – unique digital fingerprints – for violent terrorist images.

Facebook has assembled a team of dozens of low-paid contractors to not only remove terrorist propaganda but carry out in-depth investigations into suspicious users that include tracking their location and reading private messages, the Guardian has learned.

The counter-terrorism unit has special clearance to carry out investigations into user accounts if they are suspected of having links to terrorist groups identified by the US state department.

Full account access is granted to any profile once it has been flagged by algorithms looking for certain names, images or hashtags associated with terrorist groups – including Islamic State and Al-Qaida – or by user reports.

Moderators will then access the individual’s private messages, see who they are talking to and what they are saying, and view where they have been.

The team’s highest priority is to identify “traveling fighters” for Isis and Al-Qaida. Someone would be categorized as such if their profile has content that’s sympathetic to extremism and if they had, for example, visited Raqqa in Syria before traveling back to Europe. When a traveling fighter is identified – which according to one insider takes place at least once a day – the account is escalated to an internal Facebook team that decides whether to pass information to law enforcement. If a flagged account made any specific threat within a message, it would also be escalated.

Facebook did not respond to a request for comment about these investigations.

Twitter has faced a different challenge. The company has a history of fiercely defending free speech and anonymity online, where Facebook and Google have pushed for real names. Studies have shown that Twitter has been one of Isis’s preferred tools for spreading propaganda and funneling people into private messaging apps such as Telegram.

A former Twitter employee said that the company takes the problem extremely seriously, partly for legal reasons. In May, Twitter, Facebook and Google were sued by relatives of victims of the San Bernardino attack for knowingly supporting Islamic State and its extremist agenda.



Twitter has suspended thousands of accounts for promoting terrorism, but still remains a preferred tool for propaganda. Photograph: Kacper Pempel/Reuters

“They could get hit with aiding and abetting someone on the terrorist watch list,” he said.

“It’s also not good business to have people lopping off heads and using your company as part of that terrorist mechanism.”

Accounts that are clearly Isis related are taken down. And there have been a lot of takedowns: Twitter suspended 376,890 accounts for violations related to promotion of terrorism in the second half of 2016 alone.

At Google, YouTube is the platform favored by jihadists and hate preachers, although Google Docs and Google Plus are also used. Not only is YouTube hosting propaganda, but it was making money off it: earlier this year it was discovered that the ad campaigns of major brands were appearing alongside hate preachers. This led to a mass exodus of advertisers until YouTube pledged to clamp down on the problem.

The company has been experimenting with the “Redirect Method”, which delivers ads alongside search queries commonly used by people attracted to Isis. The ads link to Arabic and English channels featuring videos that counter Isis brainwashing such as testimonials from former extremists and imams denouncing Isis’s corruption of Islam.

Even with these measures in place, experts criticize the social media companies as ineffective.

“It’s all bullshit. It’s an effort to generate a veneer of corporate responsibility for the benefit of their shareholders that is frankly anything but that,” said terrorism analyst Michael Smith, who said the tech companies “understand the issues” but have “about a one-inch deep knowledge of a two-mile deep pond of centuries-old issues”.

Hany Farid, senior adviser to the Counter Extremism Project, agrees.

“They spend huge resources developing technology to find the data about users to target advertising,” he said. “I’m 100% sure that if they put even a fraction of that effort into building an early warning system they could make an impact.”

It takes minutes to find a plethora of terrorist content, including a Facebook page with a recruitment video celebrating the martyrdom of a Canadian jihadi (which Facebook deemed OK when reported). On YouTube it’s easy to find videos of English-speaking Isis fighters calling for American radicals to “stab a kafir with a knife, throw him off a building or run him over with a car.”.



Facebook page featuring a recruitment video focused on a Canadian jihadi with Isis flag in the background. When reported, Facebook did not take the page down. Photograph: Facebook

Islamic State has become increasingly sophisticated at peddling propaganda. Every communication is distributed like a multimedia ad campaign, with well-produced videos, radio shows and branded hashtags.

The content is amplified by an army of pro-Isis media groups that push it out across the internet, uploading videos hundreds of times in different locations to make it harder to police.

“Social media companies don’t understand that it’s a systemic problem across the whole online universe. That’s why the terrorists are winning,” said Eric Feinberg from cyber intelligence firm GIPEC. “They need to block communications strands, not individual accounts.”



The description of this video (translated using Google Translate) explains how to make a belt bomb with a link to more detailed instructions. Photograph: YouTube (found by Eric Feinberg)/YouTube

The agreement to come together to tackle the problem is a start. If the tech companies aren’t seen to be doing more, they may be forced to by regulators.

Europe has already proposed such a crackdown and Smith believes it won’t be more than two years until another “mass casualty” domestic attack makes the US follows suit. Regulators may want tech companies to ensure they can better identify and locate their users (a “know your customer” for social media) and more actively police their networks for terrorist propaganda – proposals civil liberties campaigners will baulk at.

The prospect of policymakers intervening may be enough to twist the arms of US-headquartered tech companies into self-regulation, Smith said.

This wouldn’t, of course, be a panacea for terrorism, but it could make it harder for extremists to find new recruits.

“If it all disappeared tomorrow, it would be a tremendous win for global security. If you pushed Islamic State into darker spaces on the web the influence and capacity to recruit and incite violence would be vastly diminished.”

Contact the author: olivia.solon@theguardian.com


comments powered by Disqus