Europe Needs to Do More than Scratch at the Ramparts of Big Tech's Castle

Gormenghast is a fantasy novel series authored by British writer Mervyn Peake around 1950. It tells the story of Castle Gormenghast, a vast, sprawling, remote, labyrinthine, decadent structure inhabited by eccentric autocrats – His Lordship and Her Ladyship – living in mad isolation with their arcane and bizarre rituals. They rule over villagers who eke out a miserable existence at the bottom of the ramparts. Once a year, the villagers – known as Bright Carvers – are allowed into the Castle to present His Lordship with their artifacts, all of which are ceremoniously set on fire. The Carvers are then ejected again to their life of alienation beyond the Outer Walls of the castle.

There is some unsettling analogy between this scenario and the predicament of Europe vis-à-vis US Big Tech. Notwithstanding the early start, Europe failed to move the dial with antitrust enforcement in digital markets – case selection has been sometimes eccentric, driven by complainants; cases have often been too narrow, focused on partial conduct and not on the big “ecosystem” picture; theories of harm have been sometimes inapt, pushing everything into the “exclusion” box and failing to engage with exploitation; remedies lacked courage and foresight.

We (as a European, I am entitled to use the pronoun) then adopted ex-ante regulation – the Digital Markets Act (DMA) and its companion, the Digital Services Act (DSA) – in a frenzy of enthusiasm and self congratulation (more Brussels effect!), and in the hope that setting out the rules of the game would induce Big Tech to play nice and comply with our European values spontaneously and good naturedly. But the reality of implementing the law has now revealed itself, and it’s clear that despite initial optimistic proclamations the law is not remotely self executing. Rather, it is every bit as hard to enforce as antitrust had turned out to be beforehand.

Regulation is necessary but…

Can we hope the DMA will eventually make a difference to European digital markets? As the law required, compliance reports were submitted last March by all designated “gatekeepers,” and a number of notable changes have been achieved (Apple allowing choice screen for the web browser, and allowing alternative app stores on iOS; Meta making WhatsApp interoperable with other messaging services; Google making some concessions on sharing of data across services). However the submission of compliance reports was followed in very short order with the opening of multiple non-compliance investigations: Apple and Google over whether their app store policies breach obligations to allow developers to direct consumers to their own offers, without charge; Apple also on whether its implementation of the choice screen may inhibit users from exercising choice; Meta for its "pay or consent" model, and Google for giving preferential treatment to its own vertical services over those of third parties. Apple’s new fee structure for its App Store and alternative app stores is also under investigation, and so is Amazon for potentially favoring its own brand products in the Amazon Store.

Of course, “this is early days” and there is much hard work ahead. Are we surprised? That regulation needs to be detailed, prescriptive, relentless? That Big Tech is not falling into line spontaneously and doing what complainants would want them to do? As if we did not understand incentives, cost/benefit analysis, time value of money and net present values. Everything that is delayed even six months is a victory for gatekeepers, as the value of preserving the status quo (however briefly) is high.

Note also that the most vocal corporate advocates of regulation are still large companies (mostly also from the US) with the resources and the perseverance to invest in pushing regulators forward – the likes of Epic, Spotify, Match, Slack, and DuckDuckGo. Smaller European digital firms are not queuing up to say this is going to be transformational for them. Regulation is not turning out different in this sense from prior efforts at European antitrust enforcement, where priorities and results have been driven by the most resourced complainants but have not assisted European startups and small businesses. And just as in antitrust, breakups and structural remedies are “out of scope” in European regulation. So, we will continue to see de facto behavioral rules pinned onto unwilling targets, which have multiple ways of throwing grit into the aspirational “compliance” machine. Malicious? No, rational.

But there’s a deeper question. What can the rules actually do even if they were implemented? Even as Big Tech publicly moans, gatekeepers privately admit the DMA’s “do and don’t” prescriptions are not transformative nor an existential challenge to their essential business model. Complying is annoying and costly, but not really threatening to the core business, nor capable of truly dispersing power. Because therein lies the problem: regulation is by definition an attempt to impose external constraints on what is there, but does not dismantle any part of The Castle.

Will this increase competition? What about autonomy (and democracy)?

It has been curious for me to see how readily European academe has developed a fandom around the DMA. It is clear the DMA provides the opportunity for endless seminars, articles, workshops and the like for academics, but it is also telling that in my experience, no one with an ounce of business savvy (aside from big hopeful complainants) shares this enthusiasm today. Because aside from the usual platitudes that “fairness and contestability” will ipso facto create competition as a result of some magical “level playing field,” it is highly dubious that any of this will truly improve the position of European businesses, boost European productivity or help unleash European innovation.

Ask small and medium sized companies in the European digital industry. They will say they have been sold the hope that the DMA would lead Big Tech to spontaneously “do the right thing” (despite 15 years of resistance), and they would be able to seamlessly interoperate, have distribution without paying a toll, be able to preserve all their traffic for themselves and establish a direct relationship with their customers. But this is always going to be a struggle. However much regulators are striving to bring this about, Big Tech remains several steps ahead and has multiple ways to deliver less than what complainants would hope. Because, in the end, they own The Castle.

This is the sense in which I find the Gormenghast analogy so compelling. For all our pioneering regulation, we are still the Bright Carvers scratching at the door of the castle and pleading “let us in! we have made new rules!” – while His Lordship and Her Ladyship look down the ramparts with splendid ennui: “Who are you? We own this.” And what is the “this” that they own? Not just distribution channels and operating systems, but a deep network of all of the essential assets: proprietary closed source code, data vaults, cloud compute engine rooms, the network of physical pipes that enable instantaneous transfer of information. The Castle has reduced the internet to a collection of proprietary access points which we Europeans do not own or even control, and on which we are profoundly dependent.

US Big Tech firms essentially control all of the infrastructure essential to operate a digital society in Europe: content discovery, communications, identity, commerce, payments, data, advertising - as so clearly diagnosed by Maria Farrell and Robin Berjon (We Need To Rewild The Internet). But where is the public policy discussion of this? Where is academe? Francesca Bria has long been the strongest voice on this in Europe (see for instance here and here); there are a few others, but the discussion of “Digital Public Infrastructure” is still fragmented and uncoordinated. All the while, Big Tech continues to bribe governments with the promise of computing power or data centers, AI toys, and “free” training for civil servants. And on it goes.

That Europe has failed to develop its own Castles reflects of course a multitude of sins on our part, from fragmentation and lack of scale at national level to the historic power struggle between telco carriers and the computer industry, to an implicit strategy of handing over our markets to US corporations in exchange for US defense dollars. But this bargain is coming to haunt us in an era of major geopolitical uncertainty. If we don’t own any of the infrastructure which is used to deliver services to Europeans, how do we ensure our autonomy and strategic independence in the face of growing risks? Ultimately, how do we protect democracy? Content moderation rules under the Digital Services Act (DSA) are not remotely enough either. What happens when (not if) the infrastructure which we rely on, but do now own, turns hostile? Services are already being withheld from European consumers on grounds the regulatory environment is too uncertain (see Meta and Apple announcing that they would not release certain AI services in Europe). And this will likely get much worse, as Henry Farrell and Abraham Newman argue in Underground Empire: How America Weaponized the World Economy and Marietje Schaake concludes in The Tech Coup: How to Save Democracy from Silicon Valley.

Even short of explicit hostility, our absurd levels of concentration and dependency in the digital ecosystem mean that any glitch affecting The Castle becomes a systemic issue that shuts all of us down (not science fiction, just see the recent CrowdStrike episode). And how do we anticipate that economic growth and innovation can truly arise locally from businesses whose ability to operate has been suppressed, and is ultimately always going to be in the gift of The Castle?

What should Europe do? Build a “Euro Stack”

As we approach priorities and plans for the new European Commission mandate for 2024-29, we have been warned: a report by former Italian Prime Minister Enrico Letta commissioned by the EC itself on European “competitiveness" and a forthcoming highly anticipated report by another former Italian Prime Minister, Mario Draghi, have both made clear the need for major new investment in infrastructure for Europe, and for a muscular industrial policy to drive this forward. We need to demand that digital infrastructure is a major part of this, at all levels of the digital stack: from apps to hardware, from open source code that enables applications on diverse operating systems, to physical infrastructure which needs to be assembled and joined together.

If we continue to look away, and tinker at the edges with regulation, vertically integrated Big Tech Castles with proprietary points of access are forever going to prevail in the acquisition and retention of data which is core to product and service offerings. They are forever going to engage in practices that protect the ramparts, from setting access terms to shaping the direction of innovation to the tools they choose to make available, the flow of news and the structure of the media ecosystem, and the development of AI products (where European efforts dwarf relative to the muscle of American “hyperscalers” with command of the cloud, data centers, chips, data, and mountains of cash).

Of course building a “Euro Stack” which is distinct, independent, and democratic is a major lift. It requires not only technical competences and the right governance rules, but also political commitment and capital. A huge task. Yet there can be no alternative for Europe but to make this commitment, once one appreciates the scale of our dependency, and that we cannot fully pry open the doors of The Castle. The digital infrastructure we rely on today is grotesquely concentrated in the hands of US giants who will not promote the growth of European business nor ensure fair data use. Today, digital infrastructure should be seen as basic infrastructure, where the state has a role to play with patient capital, and it should be thought of as a legitimate focus for industrial policy much as we think of roads, airports, electricity, and water grids.

Others are moving ahead: why not Europe?

The irony is that the developing world has made progress in this direction, whatever their motivation in their particular circumstances, be it circumnavigating US Big Tech, economic development, inclusiveness, experimenting with digital democracy, autonomy, sovereignty, or some combination of these. There are notable examples: “India Stack” is a remarkable effort in record time which started from digital identities to payments to multiple verticals (mobility, green energy, health, education), and is now moving into other parts of the stack. Taiwan is another example which includes a startling implementation of digital democracy. Brazil is a pioneer in digital payments which are independent of the established rails, and also moved further along the stack. There are others, from Singapore to various developing countries, which are exploring replicating or adopting the India Stack, among other approaches.

Europe has not been idle of course, and has made some effort on a number of “building blocks,” but we don’t have a European cloud offering. We have a few supercomputers that are mainly confined to academic research, and we are still working towards privacy-preserving digital identities independent of Big Tech (an essential step), as well as pursuing digital wallets and payment possibilities.

So what needs to happen? A major “lift”

Europe needs to lift these existing components – some national, some at the European Commission level – into a coherent structure, with a coherent funding strategy which must include investment from the European Commission and the Member States, but also private capital, philanthropy, and venture capital. The conversation needs to be about what legal instruments, what governance structures, what financing tools can be used (not just the eternal IPCEI example that gets mentioned every time as the great EC funding innovation); how do we elicit private investment to participate, and how do we embed data protection principles in products and services by design. Making progress will require European industry to be at the table, including SMEs, and for the conversation not to be co-opted by lobbying by the usual interest groups (telcos are just one example). When Europe wanted to counter Boeing’s monopoly it did get its act together and stitched together Airbus. We need a similar effort for tech.

Of course, we also do want strong regulation and stronger competition enforcement. Antitrust in Europe is what it is – we will never break up a US giant (though it is significant that the US courts might, and the recent victory of the Department of Justice and the States against Google in the Search case may create an opportunity there). Regulation might get us some of the way along, even if it will never address the underlying power problem. It’s fine to have interoperability, choice screens, additional app stores, not to use data across services, not to self preference, and all that.

But in addition to that, we need the Tiers État – the Third Estate of French Revolution glory, in practice our democratic institutions representing the people – to step in and create an alternative to The Castle, with democracy and competitive markets instead of monopoly corporate rule. I am not against corporations per se, but we cannot forever run every European bit of data and business transactions in an AWS or a Microsoft cloud – or through Google cables. We need to push forward an open and interoperable infrastructure which is ours, built in Europe using European tech, and on which we can build digital commons, develop digital European services in multiple verticals, create a more responsible information ecosystem independent of US billionaires, provide scale to innovators, and ultimately support democracy. We will not replace the US Big Tech Castle with its antiquated concentration of power, but we need a parallel “Euro Stack” which is ours.

The Gormenghast series ends with the young Lordship abandoning the decaying castle and going alone into a world which is unknown, but holds the promise of being less bleak. How can we look the other way, put all the eggs in the basket of regulation, and fail to “go it alone” in building and strengthening our capacity in the next global order? This discussion needs to accelerate now. Other countries have leapt ahead. We have a window of opportunity as we catch up with the idea that we absolutely need industrial policy to tackle our multiple structural weaknesses. Public good investment in digital infrastructure, leveraging existing disparate efforts (including flagship initiatives like the Digital Europe Programme and others that are as yet fragmented and uncoordinated) needs to be a key part of that vision.