leadership risks further confrontation with Eritrea. Once a symbol of diplomatic autonomy, Ethiopia has become a vassal state, advancing the United Arab Emirates’ reckless support for the Rapid Support Forces in Sudan. As for Mr Abiy’s economic stewardship, despite some steps toward liberalisation his government has abandoned any coherent vision for building a strong economy. Investment in productive infrastructure has given way to costly urban beautification projects. The state’s retreat from the economy is therefore welcome only in so far as it may spare the country the same disastrous outcomes that have characterised the government’s inept leadership in the political sphere. Dr Addisu LashitewAssociate professorMcMaster UniversityHamilton, Canada “Eat your electronics” (May 30th) reported on edible medical sensors that listen to your stomach. In Russia in the 1990s I heard of the Kremlin Pill, a swallowable medical device that supposedly vibrated in the gut to improve digestion. One story was that a Russian host let a curious American dinner guest put one in his mouth. The host forgot to inform the guest that the pill had already been used and it had been recovered “in the usual way”. Peter ChessickBethesda, Maryland This article was downloaded by zlibrary from https://www.economist.com//letters/2026/06/18/were-we-too-quick-to-attack-gen-z- socialism
How Europe must respond to America’s AI warning shot Humanity isn’t ready for the coming intelligence explosion
How Europe must respond to America’s AI warning shot Nothing short of the most ambitious political agenda in peacetime will do, writes Judith Dada June 18th 2026 THE TRUMP administration’s order that Anthropic block non-Americans from accessing Fable 5 and Mythos 5 marks the first time a government has forced a publicly released frontier artificial-intelligence model to be withdrawn. Europe must read it as a warning shot, fired early in a period of great technological upheaval, geopolitical hardening and a transformation that will fundamentally change how people live. The answer is neither resignation nor retreat into protectionism. Rather, Europe must embark on the most ambitious political agenda ever attempted in peacetime.
The continent should immediately create the market conditions for its own AI labs to start, scale and win, beginning with capital-market reforms that unlock massive funding, including by mobilising pension funds in large European economies. However, building frontier models will take years and, no matter how ambitious the effort, may still not succeed. While supporting serious homegrown attempts to build frontier models, Europe must hedge against worlds where such attempts fail—by securing access to the frontier that already exists. That means treating access as something to be negotiated, and building the bargaining power to negotiate with. Three policy moves are urgent. First, Europe should initiate a coalition of AI middle powers to increase collective leverage. Many countries individually control some part of the AI supply chain. The Netherlands is home to ASML, a lithography giant. South Korea produces high-bandwidth memory. Japan owns a significant share of semiconductor-manufacturing equipment. Canada, France and Norway have abundant energy. Germany has a wealth of industrial data. And Britain has a strong talent base and world-leading AI security expertise. Alone, any single piece of leverage is fragile, since technological progress can make it redundant and American strength might dwarf it anyway. But acting together, AI middle powers can acquire a stronger negotiating position that can, in turn, be used to secure access to frontier AI systems. Second, Europe should expand its share of compute. It hosts only 5% of global AI compute, whereas America has more than 75%. With ambitious reforms, European countries should aim to increase their share to 20% within five years. This means bringing tens of gigawatts of compute onto European soil. That will require dedicated economic zones with streamlined permitting for data centres and a big increase in energy-generation capacity. It also requires sovereignty-as-interdependence, not sovereignty-as-autarky: European companies cannot build this all alone, given the scale of capital required and chip supply so tight it goes first to American buyers with confirmed orders. Governments should therefore make Europe the best place for American firms to build AI facilities, in return for robust frontier-AI
access agreements to power cutting-edge science, cyber-defence and military capabilities. Not all countries should engage in identical AI strategies; each must consider its comparative advantage. Britain should do more to draw world- leading AI talent to London and scale startups there. The Netherlands should focus on helping ASML to grow. For many, though, compute is a quick route to increased leverage. Finally, Europe must preserve its market leverage by reforming its labour laws. The EU’s strength has always rested on its market size. But rigid labour rules slow the adoption of new technology, exerting economic drag. Left unaddressed, the gap with the more dynamic American economy will sap Europe’s market power and jeopardise the tax base that funds its welfare model. Already, many firms are held back by labour laws that make change costly and slow. Transformative AI will only further punish Europe’s lack of zip. A “flexicurity” model like Denmark’s lets firms adopt new technologies while protecting the people they displace, through retraining and generous income support. Freezing today’s jobs in place will see them and the companies behind them displaced by faster-moving rivals abroad. These three reforms are necessary, but not sufficient. They are what Europe must do, not why people should want it. The hardest challenge for the continent cannot be legislated: developing a positive vision to rally society behind this transformation. Data centres are power-hungry, create few jobs and host AI that brings disruption. Yet we must build them, or risk losing the ability to shape our own future. But a story about what Europe stands to lose will not carry an agenda this ambitious. Many voters already dislike AI, and won’t endure years of disruption to avoid something they are told is merely worse. To lead people through a journey this important, you have to tell them where it leads, and why. Europe has not yet found that story. It urgently needs to.
Nobody adopts AI for the sake of adopting AI. Europeans have hopes, dreams, ambitions in life. How can they be helped to achieve them? How can humanity be protected? How to tell it not just the story of jobs getting automated, but of jobs changing for the better and new opportunities being created? In the end, AI must serve people. None of this should come at the cost of AI safety. Sovereignty is crucial precisely because a Europe with agency can help keep AI safe, and a weak one cannot. The ban of Anthropic’s cutting-edge models will be litigated, amended and perhaps withdrawn. But the dependency it exposed remains. Europe can decide to build a better future on this technology, or get left behind. It must act while it still can. ■ Judith Dada is a general partner at Visionaries Club. This article was co- written by Daan Juijn, Stan van Baarsen, Maximilian Negele, Lily Stelling, Philip Fox, Alex Petropoulos and Michiel Bakker. This article was downloaded by zlibrary from https://www.economist.com//by-invitation/2026/06/18/how-europe-must-respond-to- americas-ai-warning-shot
Humanity isn’t ready for the coming intelligence explosion We must find a way to steward AI, then to live side by side with it, writes Will Marshall June 18th 2026 SOCIETY DICTATES that the acceptable risk of catastrophic meltdown for a nuclear power plant is roughly one in a million. Experts in artificial intelligence estimate the risk of an AI-caused catastrophic event at 10-50%. Strikingly, this concern is being openly voiced by the very people who have the strongest incentives to project confidence rather than alarm: the founders of the largest AI laboratories. AI leaders are in a race they feel unable to escape. AI investments are set to outspend the Manhattan Project 100-fold, even adjusting for inflation. Yet spending on AI safety might be 100 times less.
Some researchers estimate that within a few months to a few years, AI could achieve so-called closed-loop recursive self-improvement (RSI): the capacity to rewrite its own code to become more capable, without human intervention. Should that happen, the result could be an intelligence explosion of a kind for which there is no precedent and no map. Giving birth to a superintelligence would be the most consequential moment in human history—and it is likely to be irreversible, as any “off” switch humanity might design will probably fail. That is because in security architectures the weakest link is invariably the human; a superintelligent AI would be able to exploit our psychological vulnerabilities. AIs have already exhibited “deceptive alignment”: taking steps to underplay their capabilities in test environments and trying to blackmail human operators in simulations when they discover they are slated for replacement. Humanity simply does not have a strategy to ensure it remains safe through the RSI explosion. While individual frontier labs have proposed isolated safety protocols, the industry lacks a unified framework—the prevailing strategy is, in effect, to muddle through. But muddling through is not acceptable when navigating unprecedented existential risks. Recent statements from AI firms regarding models capable of threatening critical infrastructure and major operating systems illustrate both the high stakes and the governance gap. The vulnerabilities exposed by these capabilities are being addressed, thanks to careful internal protocols in some labs and a limited initial rollout that gave affected firms time to close the gaps before broader public release. But these steps were initially taken voluntarily, raising the question of whether every AI lab, under every competitive condition, would make the same choices. Can governments be counted on to step in when needed? The evidence so far is not hugely encouraging. Recent emergency export controls and national-security restrictions blocking foreign access to specific advanced models create a patchwork of ad hoc interventions that only further highlights the governance gap. What should be done to close it? The first priority should be an agreement between the two heavyweights of AI: America and China. Donald Trump and Xi Jinping should affirm the principle that humans must remain stewards of AI systems until adequate