Top storiesNew ZealandPoliticsBusinessEntertainmentSportsWorld

Jacinda Ardern's Christchurch Call: Algorithms remain in control, and secret

Thursday, 16 May 2019

Our experience on social media is largely determined by algorithms, which are powered by huge amounts of personal data.
Our experience on social media is largely determined by algorithms, which are powered by huge amounts of personal data.

Two months after a suspected white supremacist live-streamed the fatal shooting of 51 worshippers in two mosques in Christchurch, Prime Minister Jacinda Ardern has announced a pledge to eliminate terrorist and extremist content online.

The three-page, non-binding pledge, known as the 'Christchurch Call', was signed by 17 governments (as well as the European Commission) and eight technology companies, during a summit in Paris co-hosted by Ardern and French President Emmanuel Macron.

Those companies were Amazon, Facebook, Dailymotion, Google, Microsoft, Qwant, Twitter, and YouTube. (Amazon recently made headlines for being a significant source of funding for 8chan, the message board where the man accused of the Christchurch terror attack posted his manifesto.)

'This attack was part of a horrifying new trend that seems to be spreading around the world,' Ardern wrote in an opinion piece in the New York Times. 'It was designed to be broadcast on the internet.'

**READ MORE:

Christchurch Call: Tech companies, 17 governments sign New Zealand-led pledge

Facebook says new rule would have stopped Christchurch shooter livestreaming

How YouTube spreads extremist ideas that inspire events like the Christchurch mosque shootings**

Indeed, it was announced on an 8chan message board and live-streamed on Facebook . Video and photos were quickly uploaded onto other platforms and shared around the world — despite frantic efforts from some to scrub the content from their sites.

The 'Christchurch call to action', held alongside related events for a gathering of digital ministers from leading industrial nations, was a good first step, everyone seems to agree. Whether there will be further steps remains unclear.

The pledge focussed on the most violent content, deliberately avoiding a broader debate about hate speech. 'This isn't about freedom of expression, this is about preventing violent extremism and terrorism online,' Ardern said in April.

Prime Minister Jacinda Ardern in Paris following a civil society event as part of the Christchurch Call summit.
Prime Minister Jacinda Ardern in Paris following a civil society event as part of the Christchurch Call summit.

Presumably, the comments were meant to allay concerns about infringing freedom of speech. But they didn't work — the United States refused to sign the pledge.

In a statement, the White House said it felt the document could present constitutional concerns, potentially conflicting with the First Amendment. While it 'stands with the international community in condemning terrorist and violent extremist content online', the US 'is not currently in a position to join the endorsement'.

That was a bit of a blow, given there were already criticisms the so-called global effort appeared to ignore some of the world's largest platforms.

Although some social media platforms in China, for example, are as big, or bigger, than their American counterparts, Ardern didn't invite them (other than to say, when questioned, 'any country, any company' was welcome).

Henry Chung, associate professor in marketing at Massey University, says it was a missed opportunity countries such as Japan, Korea, and China weren't present, along with companies such as Line, WeChat, and Weibo.

'I think it's good we have some Western firms to start with but the governments around the world need to recognise this is only the first step and much more work needs to be done.'

Video and photos of the massacre also circulated on Asian social media platforms following the attack. There will be 'big loop holes' in global efforts without their collaboration, he says.

'Although they may not be 100 per cent on board right away, I'm pretty sure they would take some positive action over time. The most important thing is to get them involved.'

The pledge was already freighted by multiple tensions, says University of Canterbury lecturer in data science, Giulio Valentino Dalla Riva.

Tensions between multinational corporations that are suspicious of regulations, and governments that need to think about safety and inclusion. Between 'artificial intelligence will fix it' solutions, and cultural responses. Between the call for more surveillance, and the need for community- and education-based, inclusive responses. 

The governments' commitments include building more inclusive societies, ensuring effective enforcement of applicable laws, and encouraging media outlets to apply ethical standards when reporting on terrorism.

The list for online service providers is longer. It includes taking transparent measures seeking to prevent the upload of terrorist and extremist content, improving community standards or terms of service, implementing public reporting on the quantity and nature of terrorist and violent extremist content being detected and removed, and reviewing algorithms that may drive users towards terrorist and violent extremist content.

But how can measures be transparent while the algorithms that shape our experiences online are proprietary? And how can technology companies promise to act 'without compromising trade secrets'? The pledge asks governments to focus on 'resilience and inclusiveness of our societies', yet also insists on 'technical solutions'. 

While he believes the document is 'overall positive', Della Riva says we simply can't trust online service providers to apply censoring, automated decisions (such as algorithms) that are beyond public scrutiny.

'What the Christchurch Call shows is that these companies will act, under pressure, but also that they will do the least they possibly can,' writes Marianne Elliott, lead researcher of a recent report on the impact of digital media on democracy. 

Dr Kevin Veale, Massey University lecturer in media studies, agrees. 'More has happened than I expected and it's happened faster than I expected,' he says.

'Social media companies definitely want to look good and avoid external regulation. But today some very specific commitments have been made. I thought it'd be much more of a PR [public relations] thing.'

The difficulty, he adds, will be holding the companies accountable. 'This door doesn't open very often, they don't come under this sort of pressure very often, so we have to make the most of it.'

He was encouraged by New Zealand and global investors worth more than $5 trillion uniting to pressure the companies to follow through on their commitments. 

We've had 'these sorts of discussions' before, he says, but not with this sort of financial pressure.

'The optics of investors saying they don't want to invest in companies that are facilitating terrorism and extremism, that they'll go elsewhere — that will affect [the companies'] bottom line.'

Writing against a backdrop of reports about Facebook as a channel for fake news and propaganda, technology writer Charlie Warzel said the algorithms have already won; the machines are running the asylum. CEOs aren't in charge of these platforms, algorithms are.

And while they remain secret, any attempt to monitor platforms' progress is compromised.

'I'm delighted this has been a collection call to action,' Veale says. 'The hard thing will be keeping everyone accountable.'