STAND-OFFS between nations are nothing new. But a very public spat between a government and a commercial company, in which each accused the other of taking citizens hostage and threatened sanctions, certainly seemed novel when it broke out this February.
This was the case of Facebook versus Australia, in which the tech giant briefly cut off access to some parts of the web through its platform for its 17 million Australian users, in response to a proposed law that would force it to pay for linking to news stories. Opinions are still divided on the rights and wrongs – but this skirmish looks like just a foretaste of bigger battles to come.
Advertisement
Across the world, governments are concluding that tech giants such as Facebook and Google exercise too much power and are undermining the public good by allowing hate speech and misinformation to proliferate. Not only in Australia, but also in the UK, the US, the EU and elsewhere, plans are afoot to bring them to heel.
That determination brings with it risks, though. Clamp down too hard and you can damage freedom of expression, and send out the wrong signals to authoritarian regimes worldwide. Bring in different rules in different places and you risk Balkanising the internet, destroying the universality on which it is built. Not even the tech companies deny that something should be done. The question is, what?
Big tech has certainly become big. Facebook, Google and other tech companies’ incomes have ballooned as they have benefited from the changing ways we communicate and access information and services. If Facebook’s $86 billion revenue in 2020 were a nation’s GDP, for instance, it would rank 66th in the world, with an economic output roughly on a par with Sri Lanka’s.
The success of these firms is built on data about our interests and predilections gleaned from keeping our eyeballs glued to their platforms. The basic business model is advertising. The longer and more exclusively we use big tech’s sites and services to communicate, search or shop, the more the learning algorithms that underpin them find out about what we like and think – the better then to place ads we are likely to click on.
The tech companies have largely had free rein to do that. Initially, governments saw the sector as too complicated to regulate. Then it was too fast-growing a success to slow down. But that view is now changing in many parts of the world. “We’re reaching a point of maturity in these markets, and with that maturity comes the various policy debates you get to now,” says Max Beverton-Palmer at the , a think tank set up by the former UK prime minister to promote a more globalised world.
“Big tech was first seen as too complicated to regulate, then too successful to slow down”
That, plus the sense of something going seriously awry. It is often the most outrageous content that hooks us and keeps us clicking. The tech companies stand accused that their algorithms push people to peddlers of fake news and extreme content, poisoning public discourse. The algorithms also often reflect the biases of their designers, sometimes drowning out minority voices.
“The last 12 months have really demonstrated – with QAnon and the attempted insurrection at the US Capitol, and the problem of anti-vaccine conspiracy theories and disinformation around covid – that this content exists in ever-growing quantities and is influencing real-world behaviours,” says , a UK MP and former chair of the House of Commons’s Digital, Culture, Media and Sport Committee.
Facebook, the world’s biggest social network, with 2.8 billion monthly active users, often faces the brunt of criticism. Its response has been attempts at self-regulation, including introducing content moderators and last year establishing a nominally independent to review controversial decisions (see “Who watches the watchers?“). That doesn’t satisfy Collins, who is a member of the , a group of academics, politicians and civil rights activists set up to put further pressure on the company. “The model is holding people’s attention,” says Collins. “It’s not suited them to try and change the way they regulate the platform.”
“Bring in different rules in different places and you risk destroying what the web is built on”
Others agree. “The idea of disruption that’s embedded in the Silicon Valley ethos has created amazing new opportunities and tools and services that transform the world,” says Beverton-Palmer. “But when you reach a certain size with a company, there’s inertia in the system and you’re either trying to hold on to that power or you’re trying to battle against other power sources.”
, a member of the at the University of Cambridge, goes even further in her critique of big tech’s business model. “They think of it as freeing up markets and freeing up things,” she says. “I think of it as destroying or undermining existing things and rebuilding them in a way that relies on these new services.”
Perhaps surprisingly, Facebook itself doesn’t deny a need for some change. A spokesperson told Â鶹´«Ă˝: “We recognise our significant responsibilities as a company and have actively called for more government regulation.” They added: “By updating rules for the internet, we can preserve what’s best about it – the freedom for people to express themselves and for entrepreneurs to build new things – while also protecting society from broader harms.”
Detrimental effects
The fear from big tech and its defenders is that politicians trying to claw back some balance will end up tipping the scales the other way. Overly censorious laws on big tech could reduce freedom of expression, while leaving platforms so liable for the content posted on them that they can’t be effectively moderated.
So what of Facebook versus Australia? Nominally at least, the spat related to a small but unusually sensitive area of the wider concerns: how big tech’s platforms are, while amplifying those who promote hate speech and unreliable information, also undermining genuine journalism. Money that advertisers might once have paid newspapers and the like to place against their journalism, providing them with a large chunk of their revenue, now goes to the tech giants whose sites direct them to it, but who pay little or nothing towards making it. “The way private sector, laissez-faire regulation has driven technology companies is sometimes at odds with healthy discourse and has had some detrimental effect on local, independent media sources, which are also essential to a vibrant and healthy democracy,” says Lindsay Gorman at the , a US think tank.
Voluntary revenue-sharing agreements between big tech and media companies have been trialled, but increasingly governments have stepped in. In France, a “link tax” introduced this year, based on the EU’s new copyright directive, will compel Google to pay $22 million a year to French news publishers for the next three years to feature their stories in Google search results.
Google also signed a stand-alone deal on 17 February with the media business News Corp, owned by Rupert Murdoch, to stave off an Australian government law to compel them to. Facebook refused, for a time blanking access to news content from its site in Australia. The stand-off only ended after international condemnation of the firm, and some concessions from the Australian government.
While it did bring the companies to the negotiating table, Australia’s attempts to make tech platforms pay for access to news was “tremendously misguided”, says John Bergmayer, legal director at , a US-based organisation campaigning for an open internet. Some critics claimed the move was more about the Australian government protecting powerful allies in traditional media.
Others, however, are applauding an assertion of democratic rights. “The Australian case shows that when politicians decide to legislate and stick by that, the companies have to respond, even if they don’t like it,” says Collins. “I think it’s an early taste of what we’re going to see elsewhere.”
And there is much more in the pipeline. In the UK, an Online Safety Bill has been drawn up for debate this year. As currently worded, it will make UK communications regulator Ofcom responsible for ensuring digital platforms protect their users from harm – defined as anything from enabling child sexual exploitation to allowing the distribution of damaging content about eating disorders – with fines of up to 10 per cent of a company’s annual global turnover for failings.
The European Union’s General Data Protection Regulation (GDPR), in force since 2018, has already restricted how tech platforms can handle user data, and the EU is planning a similar in purpose to the UK legislation. “It will protect users and their fundamental rights, rebalancing the responsibilities of users, platforms and public authorities according to European values, placing citizens at the centre,” a European Commission spokesperson told Â鶹´«Ă˝. Individual EU countries can go further. Germany, whose history gives it a low tolerance of hate speech, recently beefed up its Network Enforcement Act, requiring digital platforms not just to remove offensive content fast or face punitive fines, but also to proactively report the worst cases to police.
Meanwhile, under a planned EU , companies found to be engaging in monopolistic behaviours will be fined up to 10 per cent of their global turnover, with the aim of levelling the playing field for smaller, nimbler start-ups. In the US, there is increasing talk across the political spectrum about breaking up the tech titans, with executives such as Facebook’s Mark Zuckerberg called to congressional hearings to explain why their companies should be spared. In India and Pakistan, governments have become more interventionist against content they feel is inappropriate being shared on social media.
“We may have to accept there’s no perfect way to rebalance the relationship with big tech”
Holistic approach
The diversity of these different approaches has some people worrying. “It’s not helpful to have too diverse a hodgepodge because you end up having to comply with the bottom line anyway,” says Gorman. She points to unintended consequences of the GDPR: following its introduction in Europe, some US news websites refused to allow European visitors to access their sites for fear of falling foul of the regulation.
Yet there is little agreement on the best way forward. Beverton-Palmer favours something like the UK’s planned online safety legislation: it is a worthy attempt to build something coherent from first principles to minimise digital harms, he says. Collins agrees, saying the proposals will target how the tech giants make money, forcing action. Cobbe is less sure. “There’s very little in the way of targeting their business models and root problems,” she says. “Some of it could be very useful, but it’s not the holistic, multipronged approach we need.”
Bergmayer favours a more direct approach, breaking up big tech where necessary. “You don’t just have one big, global telephone company for Earth, but that’s kind of where we are with Facebook,” he says. Gorman, meanwhile, thinks that at least some of the initiative for change has to come from within. “I’d love to see more initiatives at companies, whether big or small, to build a democracy by design framework into their product development, so the emphasis isn’t only on scale, scale, scale and grow, grow, grow, but also doing that in a civically minded way,” she says. That would ensure businesses have the time to think through changes to their products and the impact they could have on society, rather than letting algorithms focused on driving attention pollute the discourse, she says.
But decisions need to be made, and made quickly. “We’re at an inflection point, where either we leave things as they are so they become entrenched so deeply that it would be difficult to remove, or we intervene more strongly,” says Cobbe.
“The gold standard should be a cross-border cooperation on these ideas to set up principles and standards that will work everywhere,” says Beverton-Palmer. He also suggests that governments should bring big tech into the discussion. “We should create regulatory structures and global geopolitical agreements that incentivise things like the Facebook Oversight Board, but shore it up to make sure it is truly independent and can hold companies to account,” he says. Encouraging good behaviours alongside explicitly punishing bad ones – and creating a cohesive set of rules of the road for tech companies to follow – is likely to get us closer to rebalancing things, but we may have to accept there is no perfect solution, says Beverton-Palmer. “We’re never going to get over the problem of who watches the watchmen.”
Who watches the watchers?
Faced with the threat of increased government regulation (see main story), big tech companies often argue that self-regulation is a better answer. Yet two of the biggest tech titans, Facebook and Google, have both faced issues in recent months with internal systems of checks and balances they have set up.
Facebook’s introduction of a board of outside figures paid to about content on its Facebook or Instagram services has received a mixed response. In January, its newly constituted Oversight Board the company had taken to ban individual posts on grounds ranging from hate speech to adult nudity. It found Facebook had misstepped in four instances, and proposed nine recommendations for the company to improve its policies.
But the activist-led Real Facebook Oversight Board called the decisions “oversight theatre”, criticising the lack of transparency in how they were reached and saying they would do little to address the spread of hate speech and disinformation. “Facebook desperately needs oversight,” it said. “This is not it.”
At Google, meanwhile, the problems have centred on its artificial intelligence ethics unit. To serve up the best search results and target ads better, the company’s algorithms need to know you – but algorithms can often accentuate biases. In December 2020, Timnit Gebru, who jointly led this unit, after it asked her to withdraw a paper warning of potential biases and negative environmental impact of an AI algorithm used to learn what we search for and type, and she refused. Google claims she resigned; she claims she was sacked.
In February this year, it fired the unit’s other leader, Margaret Mitchell, who had criticised the company over Gebru’s departure, for an alleged code-of-conduct breach. It brought 10 teams within the company, including the AI ethics unit, under the control of Marian Croak, a prominent black vice-president at Google, to try to “turn around the situation”, Croak said to staff in a leaked recording.
