When historians look back on this period, one of the things that they will find remarkable is that for a quarter of a century, the governments of western democracies slept peacefully while some of the most powerful (and profitable) corporations in history emerged and grew, without let or hindrance, at exponential speeds. They will wonder at how a small number of these organisations, which came to be called “tech giants” (Alphabet, Amazon, Apple, Facebook and Microsoft), acquired, and began to wield, extraordinary powers. They logged and tracked everything we did online – every email, tweet, blog, photograph and social media post we sent, every “like” we registered, every website we visited, every Google search we made, every product we ordered online, every place we visited, which groups we belonged to and who our closest friends were. And that was just for starters. Two of these companies even invented a new variant of extractive capitalism. Whereas the standard form appropriated and plundered the Earth’s natural resources, this new “surveillance capitalism” appropriated human resources in the shape of comprehensive records of users’ behaviour, which were algorithmically translated into detailed profiles that could be sold to others. And while the activities of extractive capitalism came ultimately to threaten the planet, those of its surveillance counterpart have turned into a threat to our democracy. Some of the powers the companies wielded were relatively familiar, basically just contemporary manifestations of older kinds of industrial power: monopolistic domination of certain markets. But future historians will also note that some powers acquired by the tech giants of the early 21st century seemed genuinely novel. They included: the power to transform the public sphere by the algorithmic curation of our information feeds; the ability to silence the most powerful politician in the western world by suddenly banning him from company platforms; and the power effectively to render people invisible by delisting them from Google searches. Democracy’s long slumber ended in 2016 when two political earthquakes shook the political world – the Brexit vote in the UK and the election of Donald Trump in the US. Although both shocks were indicators of a deep malaise in liberal democracy, they were widely – but wrongly – attributed to social media. There’s no doubt that technology played a role in the upheavals of 2016, but anyone who attributes such seismic shifts just to the operations of tech companies hasn’t been paying attention to the recent history of either capitalism or democracy. In fact, blaming tech provides a convenient way of ignoring the deeper causes of the turbulence. Despite that, the focus of media and public attention has largely been on the power and role of tech companies in our societies. The years since 2016 have seen flurries of activity – antitrust lawsuits; Senator Elizabeth Warren’s presidential campaign; congressional hearings; a major investigation by the US House of Representatives; leaks from inside the companies; sensational media revelations (the Cambridge Analytica scandal, Facebook’s role in facilitating genocide in Myanmar, YouTube’s role in radicalising mass shooters etc); probes by competition authorities in the UK, the EU and elsewhere. By some counts, there are at least 70 such actions under way around the globe at the moment. In the US, for example, nearly 40 states have launched competition lawsuits against Google and the Department of Justice is pursuing one against Facebook. In Europe, the European Commission has filed competition and other charges against Amazon and Google, while a number of other tech companies have been suing Apple over its alleged anti-competitive behaviour in the management of its app store. To date, however, we’ve seen little in the way of tangible, effective curbs on tech power. Sensational media revelations or political slogans such as “Break Them Up” may create headlines and engender fevered discussion, but they are not a substitute for radical regulatory intervention or legislative action. And although US congressional hearings have improved of late, they have too often just been yelling matches in which grandstanding legislators summon tech executives for castigation. That’s not to say that there haven’t been some serious interventions by various authorities. Whopping fines for corporate transgressions have been levied on Facebook and Google, for example. In Google’s case, the European Commission has imposed a total of $9.5bn in fines on the company since 2017. The problem is that there’s little evidence that even such massive penalties constitute a serious deterrent for such insanely profitable companies. To give just one example. In 2012, Facebook was subjected to a consent decree by the US regulator, the Federal Trade Commission (FTC), in which it undertook always to obtain its users’ consent before sharing their information beyond established privacy settings. After the Cambridge Analytica scandal, the FTC ruled that the company had violated that decree and fined it $5bn, the biggest penalty it had ever levied. And the immediate result of this news? Facebook’s share price went up – from $201 to $205! The task of bringing these kinds of corporations under public control is a truly gargantuan one. As anyone who has worked in government knows, industrial regulation is hard for liberal democracies. First, it requires political will, which in turn requires public concern and popular support. Second, it needs vision and new ideas about how to extract the benefits of technology for society while minimising the harm that comes from the unrestrained corporate power that controls it. And third, it requires legislative determination and staying power, because structural change in a democracy takes a long time. All of these basic requirements have been absent in the decades when the tech giants were growing into their present dominance. Which means that democracies are now playing catch-up and, at worst, chasing horses that have long since bolted. In recent times, western governments have belatedly become converts to the idea that “something must be done” about tech power. Whether they understand the nature and scale of the task is debatable. To those who are sceptical about governments’ capacity to bring about structural change, the stock riposte is that since democracies have dealt with this kind of challenge before, they can do so again. After all, in the closing decade of the 19th century and the early years of the 20th the American republic took on the great industrial trusts assembled by the Rockefellers, Morgans, Carnegies and Vanderbilts and brought them under some kind of democratic control. But this was only possible because there was widespread public concern about the trusts’ manifest abuses of their consolidated power, concern that had been stoked by a formidable amount of investigative reporting by writers and journalists such as Ida Tarbell. This public concern was transmuted into political pressure. Three of the four candidates in the 1912 US presidential election, for example, ran on platforms that were deeply hostile to such accumulated industrial power. In contemporary democracies, however, no political party campaigns on a platform like that, for the simple reason that voters don’t seem to be all that interested in tech power. That’s not entirely surprising: public understanding of digital technology is limited by its formidable complexity. More importantly, because the internet and the services that run on it have become intimately interwoven with people’s everyday lives, they have become dependent on it, a dependence that has been vividly underscored by the pandemic. So although opinion polls may report that people are concerned about tech power, their behaviour tells a different story – that they suffer from what psychologists call “cognitive dissonance”: the stress that comes from continuing to do something that is contrary to what you believe to be right. This is the source of the “privacy paradox” that grips social media users, who fear (rightly) that their privacy is undermined by the services, but nevertheless continue to use them. High-minded disdain for such contradictory behaviour is unfair and counterproductive, because it ignores the power of the network effects that keep people locked into online platforms. Try telling a grandmother who uses Facebook to keep in regular touch with her grandchildren in Australia that her concerns about privacy are hypocritical. What critics of social media conveniently overlook is how much “ordinary” people value these “free” services, even as they may harbour suspicions about the ethics of the companies that provide them. Politicians in liberal democracies, with their gaze permanently fixed on electoral cycles, know this only too well. The Australian government had a sharp reminder of it last February when Facebook blocked news to its users in the country amid a dispute over a proposed law that would force it and Google to pay news publishers for content. After a few days, a settlement was reached, involving changes to the proposed legislation. This was predictably followed by disagreement about who blinked first – Facebook or the government? The inescapable conclusion, though, was that the democratically elected prime minister who would ban access to, say, Instagram (a Facebook property) has yet to be born. Sanctimonious criticism of social media users’ lack of moral fibre is unfair also because it attributes to them more agency than they actually possess. Most people imagine that if they decide to stop using Gmail or Microsoft Outlook or never buy another book from Amazon then they have liberated themselves from the tentacles of these giants. But the penetration and connectedness of networked technology is such that the only way of avoiding the tentacles of tech power would be to go completely off-grid. Three years ago, an intrepid journalist named Kashmir Hill conducted an interesting experiment to see if she could avoid using Amazon, Facebook, Google, Microsoft or Apple services. “Over six weeks,” she reported, “I cut them out of my own life and tried to prevent them from knowing about me or monetising me in any way – not just by putting my iPhone in a drawer for a week or only buying local, but by really, truly blocking these companies from accessing me and vice versa. I wanted to find out how hard it would be – or if I could even do it – given that these tech giants dominate the internet in so many invisible ways that it’s hard to even know them all.” The results of Ms Kashmir’s experiment were fascinating. She demonstrated that our lives now run on a technical infrastructure that is owned, operated and controlled by a handful of giant corporations, from which there is currently no escape unless you plan to hibernate. But perhaps the most sobering outcome of the experiment was the extent to which almost every digital service we use is underpinned, in one way or another, by Amazon’s cloud-computing service, AWS. And in a way this may help to explain why western governments are so chary of taking on tech giants, particularly Amazon. For it turns out that even the security services of major democracies are using Amazon Web Services (AWS). Just to take a couple of examples, the CIA has been using it since 2014 and recently it was revealed that the UK’s spy agencies have given a £500m-plus contract to AWS to host classified material to boost the use of data analytics and “AI”. GCHQ led the procurement of this high-security cloud system, which will also be used by MI5 and MI6 as well as the Ministry of Defence. Of course, these arrangements are accompanied by the usual soothing official bromides – alles ist in Ordnung and so on. But it does make one wonder how keen a future British government might be to impose stringent competition regulations on its new partner in national security. The USP of liberal democracies is that they are governed by the rule of law. But legal climates change over time and so it has been with judicial attitudes to monopoly power over the decades. The first anti-monopoly statute was the Sherman Anti-Trust Act, passed by the US Congress in 1890. The act was crafted to prevent the concentration of power into the hands of a few large enterprises to the disadvantage of smaller enterprises. And it gave the US Department of Justice the authority to take action against offenders. Crudely put, in the view of the act, “big equalled bad” and this shaped competition enforcement and thinking over the next half century. But, as with most major statutes, consistent action proved increasingly difficult over succeeding decades as new industries evolved and circumstances changed. Then, in 1978, came a landmark book, The Antitrust Paradox, by a distinguished American jurist, Robert Bork. Bork provided a scathing critique of how the Sherman Act had become dysfunctional: originally aimed at protecting competition, it had increasingly been used to protect weak and uneconomic competitors, a perverse outcome for the US economy. Instead of focusing on corporate consolidation (ie size), Bork proposed that the prime focus of antitrust action should be consumer welfare, which in practice meant protection from unfairly high prices. The fact that over a period a corporation had grown very large was not in itself problematic, so long as there was no evidence that it was harmful to consumers. Big no longer automatically meant bad. What no one could have known in 1978 was that Bork’s view would provide a get-out-of-jail card for tech firms that grew into giants but could not be accused of harming consumers by raising prices, because their products were “free” (Google, Facebook, Twitter) or super-competitive (Amazon). The freedom that this gave to tech companies was memorably articulated by the Silicon Valley billionaire Peter Thiel in his Zero to One manifesto: “Monopoly is the condition of every successful business,” he wrote, and “competition is for losers”. The conventional wisdom embodied in The Antitrust Paradox may have explained the somnolence of democratic regulators when the companies were expanding. At any rate, it could have reduced their appetite for action at a time when it might have been more effective. In that sense, perhaps the most significant development of the past few years came in 2017 when a young graduate student named Lina Khan published a remarkable article in the Yale Law Journal. In a way, its title – Amazon’s Antitrust Paradox – with its echo of Bork’s landmark book, should have given the game away, because it mounted a head-on challenge to the conventional wisdom that regulation should focus on consumer welfare. Khan’s argument was that a company shouldn’t get a free pass just because it makes its customers happy. Benefiting from the slumber of regulators as it grew, Amazon had amassed structural power over increasing parts of the economy. It had unparalleled amounts of data on its customers, was commercially very aggressive and its massive logistical and warehousing infrastructure enabled it to wield power greatly disproportionate to its actual market share. In that sense, it had come to resemble the railroads that John D Rockefeller and his fellow titans controlled in the 1890s. “The thousands of retailers and independent businesses that must ride Amazon’s rails to reach market,” Khan wrote, “are increasingly dependent on their biggest competitor.” Just like the bad old days in fact. Her article garnered more than 140,000 hits, which made it “a runaway bestseller in the world of legal treatises”. The question was, would it, like Bork’s book four decades earlier, change the conventional wisdom on antitrust? Early indications are that it has. Khan was one of the leading figures who guided the investigation into monopoly tech power conducted by the US House of Representatives’ antitrust subcommittee. Then in November 2020 Joe Biden was elected president and in March 2021 he made Khan the chair of the Federal Trade Commission, the federal agency whose principal mission is the enforcement of civil US competition law and the promotion of consumer protection. Other indications of the change in the regulatory climate came with Biden’s appointment to powerful positions of tech critics such as the Columbia legal scholar Tim Wu and Meredith Whittaker, the woman who organised the 2018 walkout at Google, when about 20,000 employees protested against how the company handled alleged sexual harassment. These are important changes because the tech giants are all US companies and the federal government is the only public authority that can make deep structural changes in the industry. Other jurisdictions, most importantly the EU, can force changes in the way the companies operate on their territories, but only the US government can make Google divest itself of YouTube or force Facebook to set Instagram and WhatsApp free. These changes in the Biden administration are good news because they suggest that the slumbering democratic giant has finally woken up. But they’re only the beginning of what could be a long process. The last big competition case in the US, when the government tried to have Microsoft broken up for abusing its monopoly power over the PC operating system, took more than four years from launch to conclusion and ultimately failed. There’s little in the current plethora of analogous suits and actions to suggest that they will be any quicker to produce results. Years ago in his book The Confidence Trap, the political theorist David Runciman pointed out that democracies are congenitally complacent, hooked as they are on the dangerous belief that – given enough time – they can muddle through just about anything. With the climate crisis, the costs of that complacency are now becoming clear. The existential question for liberal democracies is whether that also holds for curbing the power of big tech.