It seems impossible to have a conversation today without artificial intelligence (AI) playing some role, demonstrating the massive power of the technology. It has the potential to impact every part of business, and European policymakers are on board. In February 2025, Ursula von der Leyen, the European Commission president, said, “We want Europe to be one of the leading AI continents … AI can help us boost our competitiveness, protect our security, shore up public health, and make access to knowledge and information more democratic.” Research from Nokia suggests that businesses share this enthusiasm and ambition: 84 percent of more than 1,000 respondents said AI features in the growth strategy of their organization, while 62 percent are directing at least 20 percent of ICT capex budgets toward the technology. However, the equation is not yet balanced. Three-quarters of survey respondents state that current telecom infrastructure limits the ability to deliver on those ambitions. Meanwhile, 45 percent suggest these limitations would delay, constrain or entirely limit investments. There is clearly a disconnect between the ambition and the ability to deliver. At present, Europe lags the United States and parts of Asia in areas such as network deployment, related investment levels and scale. > If AI does not reach its full potential, EU competitiveness will suffer, > economic growth will have a ceiling, the creation of new jobs will have a > limit and consumers will not see the benefits. What we must remember primarily is that AI does not happen without advanced, trusted and future-proofed networks. Infrastructure is not a ‘nice to have’ it is a fundamental part. Simply put, today’s networks in Europe require more investments to power the AI dream we all have. If AI does not reach its full potential, EU competitiveness will suffer, economic growth will have a ceiling, the creation of new jobs will have a limit and consumers will not see the benefits. When we asked businesses about the challenge of meeting AI demands during our research, the lack of adequate connectivity infrastructure was the fourth common answer out of 15 potential options. Our telecom connectivity regulatory approach must be more closely aligned with the goal of fostering AI. That means progressing toward a genuine telecom single market, adopting a novel approach to competition policy to allow market consolidation to lead to more investments, and ensuring connectivity is always secure and trusted. Supporting more investments in next-generation networks through consolidation AI places heavy demands on networks. It requires low latency, high bandwidth and reliability, and efficient traffic management. To deliver this, Europe needs to accelerate investment in 5G standalone, fiber to enterprises, edge data centers and IP-optical backbone networks optimized for AI. > As industry voices such as Nokia have emphasized, the networks that power AI > must themselves make greater use of automation and AI. Consolidation (i.e. reducing the number of telecom operators within the national telecom markets of EU member states) is part of the solution. Consolidation will allow operators to achieve economies of scale and improve operating efficiency, therefore encouraging investment and catalyzing innovation. As industry voices such as Nokia have emphasized, the networks that power AI must themselves make greater use of automation and AI. Policy support should therefore extend to both network innovation and deployment. Trust: A precondition for AI adoption Intellectual property (IP) theft is a threat to Europe’s industrial future and only trusted technology should be used in core functions, systems and sectors (such as energy, transport and defense). In this context, the underlying connectivity should always be secure and trusted. The 5G Security Toolbox, restricting untrusted technology, should therefore be extended to all telecom technologies (including fiber, optics and IP) and made compulsory in all EU member states. European governments must make protecting their industries and citizens a high priority. Completing the digital single market Although the single market is one of Europe’s defining projects, the reality in telecoms — a key part of the digital single market — is still fragmented. As an example, different spectrum policies create barriers across borders and can limit network roll outs. Levers on top of advanced connectivity To enable the AI ecosystem in Europe, there are several different enabling levers European policymakers should advance on top of fostering advanced and trusted connectivity: * The availability of compute infrastructure. The AI Continent Action Plan, as well as the IPCEI Compute Infrastructure Continuum, and the European High-Performance Computing Joint Undertaking should facilitate building AI data centers in Europe.   * Leadership in edge computing. There should also be clear support for securing Europe’s access to and leadership in edge solutions and building out edge capacity. Edge solutions increase processing speeds and are important for enabling AI adoption, while also creating a catalyst for economic growth. With the right data center capacity and edge compute capabilities available, European businesses can meet the new requirements of AI use cases.  * Harmonization of rules. There are currently implications for AI in several policy areas, including the AI Act, GDPR, Data Act, cybersecurity laws and sector-specific regulations. This creates confusion, whereas AI requires clarity. Simplification and harmonization of these regulations should be pursued.  * AI Act implementation and simplification. There are concerns about the implementation of the AI Act. The standards for high-risk AI may not be available before the obligations of the AI act enter into force, hampering business ambitions due to legal uncertainty. The application date of the AI Act’s provisions on high-risk AI should be postponed by two years to align with the development of standards. There needs to be greater clarity on definitions and simplification measures should be pursued across the entire ecosystem. Policies must be simple enough to follow, otherwise adoption may falter. Policy needs to act as an enabler, not a barrier to innovation.  * Upskilling and new skills. AI will require new skills of employees and users, as well as creating entirely new career paths. Europe needs to prepare for this new world.  If Europe can deliver on these priorities, the benefits will be tangible: improved services, stronger industries, increased competitiveness and higher economic growth. AI will deliver to those who best prepare themselves. We must act now with the urgency and consistency that the moment demands. -------------------------------------------------------------------------------- Author biography: Marc Vancoppenolle is leading the geopolitical and government relations EU and Europe function at Nokia. He and his team are working with institutions and stakeholders in Europe to create a favorable political and regulatory environment fostering broadband investments and cross sectoral digitalization at large. Vancoppenolle has over 30 years of experience in the telecommunication industry. He joined Alcatel in 1991, and then Alcatel-Lucent, where he took various international and worldwide technical, commercial, marketing, communication and government affairs leadership roles. Vancoppenolle is a Belgian and French national. He holds a Master of Science, with a specialization in telecommunication, from the University of Leuven complemented with marketing studies from the University of Antwerp. He is a member of the DIGITALEUROPE Executive Board, Associate to Nokia’s CEO at the ERT (European Round Table for Industry), and advisor to FITCE Belgium (Forum for ICT & Media professionals). He has been vice-chair of the BUSINESSEUROPE Digital Economy Taskforce as well as a member of the board of IICB (Innovation & Incubation Center Brussels).

Meta named former Trump adviser Dina Powell McCormick to serve as president and vice chair Monday, further cementing the company’s growing ties to Republicans and President Donald Trump’s White House. In addition to a long career on Wall Street, Powell McCormick served as Trump’s deputy national security adviser during his first term. She was also a member of the George W. Bush administration. She first joined Meta’s board last April, part of a broader play by the social media and artificial intelligence giant to hire Republicans following Trump’s election. In a statement, Meta CEO Mark Zuckerberg praised Powell McCormick’s “experience at the highest levels of global finance, combined with her deep relationships around the world, [which] makes her uniquely suited to help Meta manage this next phase of growth.” Rightward trend: Powell McCormick’s time in global finance — she spent 16 years as a partner at Goldman Sachs and was most recently a top executive at banking company BDT & MSD Partners — could be a major asset to Meta as it raises hundreds of billions of dollars to build out data centers and other AI-related infrastructure. But her GOP pedigree and proximity to Trump likely played a significant role in her hiring as well. Since Trump’s election, Meta has worked to curry favor with Republicans in the White House and on Capitol Hill. The company elevated former GOP official Joel Kaplan to serve as global affairs lead last January, simultaneously tapping Kevin Martin, a former Republican chair of the Federal Communications Commission, as his No. 2. Under pressure from Republicans, last year Meta also rolled back many of its former rules related to content moderation. In 2024, the company apologized to congressional Republicans — specifically Rep. Jim Jordan (R-Ohio), chair of the House Judiciary Committee — for removing content that contained disinformation about the Covid-19 pandemic. A Meta spokesperson declined to comment when asked whether Powell McCormick’s ties to Trump and Republicans played a role in her hiring. Trump thumbs up: In a Truth Social post Monday, Trump congratulated Powell McCormick and said Zuckerberg made a “great choice.” The president called her “a fantastic, and very talented, person, who served the Trump Administration with strength and distinction!”

Venture capitalist Finn Murphy believes world leaders could soon resort to deflecting sunlight into space if the Earth gets unbearably hot. That’s why he’s invested more than $1 million in Stardust Solutions, a leading solar geoengineering firm that’s developing a system to reduce warming by enveloping the globe in reflective particles. Murphy isn’t rooting for climate catastrophe. But with global temperatures soaring and the political will to limit climate change waning, Stardust “can be worth tens of billions of dollars,” he said. “It would be definitely better if we lost all our money and this wasn’t necessary,” said Murphy, the 33-year-old founder of Nebular, a New York investment fund named for a vast cloud of space dust and gas. Murphy is among a new wave of investors who are putting millions of dollars into emerging companies that aim to limit the amount of sunlight reaching the Earth — while also potentially destabilizing weather patterns, food supplies and global politics. He has a degree in mathematics and mechanical engineering and views global warming not just as a human and political tragedy, but as a technical challenge with profitable solutions. Solar geoengineering investors are generally young, pragmatic and imaginative — and willing to lean into the adventurous side of venture capitalism. They often shrug off the concerns of scientists who argue it’s inherently risky to fund the development of potentially dangerous technologies through wealthy investors who could only profit if the planet-cooling systems are deployed. “If the technology works and the outcomes are positive without really catastrophic downstream impacts, these are trillion-dollar market opportunities,” said Evan Caron, a co-founder of the energy-focused venture firm Montauk Capital. “So it’s a no-brainer for an investor to take a shot at some of these.” More than 50 financial firms, wealthy individuals and government agencies have collectively provided more than $115.8 million to nine startups whose technology could be used to limit sunlight, according to interviews with VCs, tech company founders and analysts, as well as private investment data analyzed by POLITICO’s E&E News. That pool of funders includes Silicon Valley’s Sequoia Capital, one of the world’s largest venture capital firms, and four other investment groups that have more than $1 billion of assets under management. Of the total amount invested in the geoengineering sector, $75 million went to Stardust, or nearly 65 percent. The U.S.-Israeli startup is developing reflective particles and the means to spray and monitor them in the stratosphere, some 11 miles above the planet’s surface. At least three other climate-intervention companies have also raked in at least $5 million. The cash infusion is a bet on planet-cooling technologies that many political leaders, investors and environmentalists still consider taboo. In addition to having unknown side effects, solar geoengineering could expose the planet to what scientists call “termination shock,” a scenario in which global temperatures soar if the cooling technologies fail or are suddenly abandoned. Still, the funding surge for geoengineering companies pales in comparison to the billions of dollars being put toward artificial intelligence. OpenAI, the maker of ChatGPT, has raised $62.5 billion in 2025 alone, according to investment data compiled by PitchBook. The investment pool for solar geoengineering startups is relatively shallow in part because governments haven’t determined how they would regulate the technology — something Stardust is lobbying to change. As a result, the emerging sector is seen as too speculative for most venture capital firms, according to Kim Zou, the CEO of Sightline Climate, a market intelligence firm. VCs mostly work on behalf of wealthy individuals, as well as pension funds, university endowments and other institutional investors. “It’s still quite a niche set of investors that are even thinking about or looking at the geoengineering space,” Zou said. “The climate tech and energy tech investors we speak to still don’t really see there being an investable opportunity there, primarily because there’s no commercial market for it today.” AEROSOLS IN THE STRATOSPHERE Stardust and its investors are banking on signing contracts with one or more governments that could deploy its solar geoengineering system as soon as the end of the decade. Those investors include Lowercarbon Capital, a climate-focused firm co-founded by billionaire VC Chris Sacca, and Exor, the holding company of an Italian industrial dynasty and perhaps the most mainstream investment group to back a sunlight reflection startup. Even Stardust’s supporters acknowledge that the company is far from a sure bet. “It’s unique in that there is not currently demand for this solution,” said Murphy, whose firm is also supporting out-there startups seeking to build robots and data centers in space. “You have to go and create the product in order to potentially facilitate the demand.” Lowercarbon partner Ryan Orbuch said the firm would see a return on its Stardust investment only “in the context of an actual customer who can actually back many years of stable, safe deployment.” Exor, another Stardust investor, didn’t respond to a request for comment. Other startups are trying to develop commercial markets for solar geoengineering. Make Sunsets, a company funded by billionaire VC Tim Draper, releases sulfate-filled weather balloons that pop when they reach the stratosphere. It sells cooling credits to individuals and corporations based on the theory that the sulfates can reliably reduce warming. There are questions, however, about the science and economics underpinning the credit system of Make Sunsets, according to the investment bank Jeffries. “A cooling credit market is unlikely to be viable,” the bank said in a May 2024 note to clients. That’s because the temperature reductions produced by sulfate aerosols vary by altitude, location and season, the note explained. And the warming impacts of carbon dioxide emissions last decades — much longer than any cooling that would be created from a balloon’s worth of sulfate. Make Sunsets didn’t respond to a request for comment. The company has previously attracted the attention of regulators in the U.S. and Mexico, who have claimed it began operating without the necessary government approvals. Draper Associates says on its website that it’s “shaping a future where the impossible becomes everyday reality.” The firm has previously backed successful consumer tech firms like Tesla, Skype and Hotmail. “It is getting hotter in the Summer everywhere,” Tim Draper said in an email. “We should be encouraging every solution. I love this team, and the science works.” THE NEXT FRONTIER One startup is pursuing space-based solar geoengineering. EarthGuard is attempting to build a series of large sunlight deflectors that would be positioned between the sun and the planet, some 932,000 miles from the Earth. The company did not respond to emailed questions. Other space companies are considering geoengineering as a side project. That includes Gama, a French startup that’s designing massive solar sails that could be used for deep space travel or as a planetary sunshade, and Ethos Space, a Los Angeles company with plans to industrialize the moon. Both companies are part of an informal research network established by the Planetary Sunshade Foundation, a nonprofit advocating for the development of a trillion-dollar parasol for the globe. The network mainly brings together collaborators on the sidelines of space industry conferences, according to Gama CEO Andrew Nutter. “We’re willing to contribute something if we realize it’s genuinely necessary and it’s a better solution than other solutions” to the climate challenge, Nutter said of the space shade concept. “But our business model does not depend on it. If you have dollar signs hanging next to something, that can bias your decisions on what’s best for the planet.” Nutter said Gama has raised about $5 million since he co-founded the company in 2020. Its investors include Possible Ventures, a German VC firm that’s also financing a nuclear fusion startup and says on its website that the firm is “relentlessly optimistic — choosing to focus on the possibilities rather than obsess over the risks.” Possible Ventures did not respond to a request for comment. Sequoia-backed Reflect Orbital is another space startup that’s exploring solar geoengineering as a potential moneymaker. The company based near Los Angeles is developing a network of satellite mirrors that would direct sunlight down to the Earth at night for lighting industrial sites or, eventually, producing solar energy. Its space mirrors, if oriented differently, could also be used for limiting the amount of sun rays that reach the planet. “It’s not so much a technological limitation as much as what has the highest, best impact. It’s more of a business decision,” said Ally Stone, Reflect Orbital’s chief strategy officer. “It’s a matter of looking at each satellite as an opportunity and whether, when it’s over a specific geography, that makes more sense to reflect sunlight towards or away from the Earth.” Reflect Orbital has raised nearly $28.7 million from investors including Lux Capital, a firm that touts its efforts to “turn sci-fi into sci-fact” and has invested in the autonomous defense systems companies Anduril and Saildrone.” Sequoia and Lux didn’t respond to requests for comment. The startup hopes to send its first satellite into space next summer, according to Stone. SpaceX CEO Elon Musk, whose aerospace company already has an estimated fleet of more than 8,800 internet satellites in orbit, has also suggested using the circling network to limit sunlight. “A large solar-powered AI satellite constellation would be able to prevent global warming by making tiny adjustments in how much solar energy reached Earth,” Musk wrote on X last month. Neither he nor SpaceX responded to an emailed request for comment. DON’T CALL IT GEOENGINEERING Other sunlight-reflecting startups are entering the market — even if they’d rather not be seen as solar geoengineering companies. Arctic Reflections is a two-year-old company that wants to reduce global warming by increasing Arctic sea ice, which doesn’t absorb as much heat as open water. The Dutch startup hasn’t yet pursued outside investors. “We see this not necessarily as geo-engineering, but rather as climate adaptation,” CEO Fonger Ypma said in an email. “Just like in reforestation projects, people help nature in growing trees, our idea is that we would help nature in growing ice.” The main funder of Arctic Reflections is the British government’s independent Advanced Research and Invention Agency. In May, ARIA awarded $4.41 million to the company — more than four times what it had raised to that point. Another startup backed by ARIA is Voltitude, which is developing micro balloons to monitor geoengineering from the stratosphere. The U.K.-based company didn’t respond to a request for comment. Altogether, the British agency is supporting 22 geoengineering projects, only a handful of which involve startups. “ARIA is only funding fundamental research through this programme, and has not taken an equity stake in any geoengineering companies,” said Mark Symes, a program director at the agency. It also requires that all research it supports “must be published, including those that rule out approaches by showing they are unsafe or unworkable.” Sunscreen is a new startup that is trying to limit sunlight in localized areas. It was founded earlier this year by Stanford University graduate student Solomon Kim. “We are pioneering the use of targeted, precision interventions to mitigate the destructive impacts of heatwave on critical United States infrastructure,” Kim said in an email. But he was emphatic that “we are not geoengineering” since the cooling impacts it’s pursuing are not large scale. Kim declined to say how much had been raised by Sunscreen and from what sources. As climate change and its impacts continue to worsen, Zou of Sightline Climate expects more investors to consider solar geoengineering startups, including deep-pocketed firms and corporations interested in the technology. Without their help, the startups might not be able to develop their planet-cooling systems. “People are feeling like, well wait a second, our backs are kind of starting to get against the wall. Time is ticking, we’re not really making a ton of progress” on decarbonization, she said. “So I do think there’s a lot more questions getting asked right now in the climate tech and venture community around understanding it,” Zou said of solar geoengineering. “Some of these companies and startups and venture deals are also starting to bring more light into the space.” Karl Mathiesen contributed reporting.

Europe’s security does not depend solely on our physical borders and their defense. It rests on something far less visible, and far more sensitive: the digital networks that keep our societies, economies and democracies functioning every second of the day. > Without resilient networks, the daily workings of Europe would grind to a > halt, and so too would any attempt to build meaningful defense readiness. A recent study by Copenhagen Economics confirms that telecom operators have become the first line of defense in Europe’s security architecture. Their networks power essential services ranging from emergency communications and cross-border healthcare to energy systems, financial markets, transport and, increasingly, Europe’s defense capabilities. Without resilient networks, the daily workings of Europe would grind to a halt, and so too would any attempt to build meaningful defense readiness. This reality forces us to confront an uncomfortable truth: Europe cannot build credible defense capabilities on top of an economically strained, structurally fragmented telecom sector. Yet this is precisely the risk today. A threat landscape outpacing Europe’s defenses The challenges facing Europe are evolving faster than our political and regulatory systems can respond. In 2023 alone, ENISA recorded 188 major incidents, causing 1.7 billion lost user-hours, the equivalent of taking entire cities offline. While operators have strengthened their systems and outage times fell by more than half in 2024 compared with the previous year, despite a growing number of incidents, the direction of travel remains clear: cyberattacks are more sophisticated, supply chains more vulnerable and climate-related physical disruptions more frequent. Hybrid threats increasingly target civilian digital infrastructure as a way to weaken states. Telecom networks, once considered as technical utilities, have become a strategic asset essential to Europe’s stability. > Europe cannot deploy cross-border defense capabilities without resilient, > pan-European digital infrastructure. Nor can it guarantee NATO > interoperability with 27 national markets, divergent rules and dozens of > sub-scale operators unable to invest at continental scale. Our allies recognize this. NATO recently encouraged members to spend up to 1.5 percent of their GDP on protecting critical infrastructure. Secretary General Mark Rutte also urged investment in cyber defense, AI, and cloud technologies, highlighting the military benefits of cloud scalability and edge computing – all of which rely on high-quality, resilient networks. This is a clear political signal that telecom security is not merely an operational matter but a geopolitical priority. The link between telecoms and defense is deeper than many realize. As also explained in the recent Arel report, Much More than a Network, modern defense capabilities rely largely on civilian telecom networks. Strong fiber backbones, advanced 5G and future 6G systems, resilient cloud and edge computing, satellite connectivity, and data centers form the nervous system of military logistics, intelligence and surveillance. Europe cannot deploy cross-border defense capabilities without resilient, pan-European digital infrastructure. Nor can it guarantee NATO interoperability with 27 national markets, divergent rules and dozens of sub-scale operators unable to invest at continental scale. Fragmentation has become one of Europe’s greatest strategic vulnerabilities. The reform Europe needs: An investment boost for digital networks At the same time, Europe expects networks to become more resilient, more redundant, less dependent on foreign technology and more capable of supporting defense-grade applications. Security and resilience are not side tasks for telecom operators, they are baked into everything they do. From procurement and infrastructure design to daily operations, operators treat these efforts as core principles shaping how networks are built, run and protected. Therefore, as the Copenhagen Economics study shows, the level of protection Europe now requires will demand substantial additional capital. > It is unrealistic to expect world-class, defense-ready infrastructure to > emerge from a model that has become structurally unsustainable. This is the right ambition, but the economic model underpinning the sector does not match these expectations. Due to fragmentation and over-regulation, Europe’s telecom market invests less per capita than global peers, generates roughly half the return on capital of operators in the United States and faces rising costs linked to expanding security obligations. It is unrealistic to expect world-class, defense-ready infrastructure to emerge from a model that has become structurally unsustainable. A shift in policy priorities is therefore essential. Europe must place investment in security and resilience at the center of its political agenda. Policy must allow this reality to be reflected in merger assessments, reduce overlapping security rules and provide public support where the public interest exceeds commercial considerations. This is not state aid; it is strategic social responsibility. Completing the single market for telecommunications is central to this agenda. A fragmented market cannot produce the secure, interoperable, large-scale solutions required for modern defense. The Digital Networks Act must simplify and harmonize rules across the EU, supported by a streamlined governance that distinguishes between domestic matters and cross-border strategic issues. Spectrum policy must also move beyond national silos, allowing Europe to avoid conflicts with NATO over key bands and enabling coherent next-generation deployments. Telecom policy nowadays is also defense policy. When we measure investment gaps in digital network deployment, we still tend to measure simple access to 5G and fiber. However, we should start considering that — if security, resilience and defense-readiness are to be taken into account — the investment gap is much higher that the €200 billion already estimated by the European Commission. Europe’s strategic choice The momentum for stronger European defense is real — but momentum fades if it is not seized. If Europe fails to modernize and secure its telecom infrastructure now, it risks entering the next decade with a weakened industrial base, chronic underinvestment, dependence on non-EU technologies and networks unable to support advanced defense applications. In that scenario, Europe’s democratic resilience would erode in parallel with its economic competitiveness, leaving the continent more exposed to geopolitical pressure and technological dependency. > If Europe fails to modernize and secure its telecom infrastructure now, it > risks entering the next decade with a weakened industrial base, chronic > underinvestment, dependence on non-EU technologies and networks unable to > support advanced defense applications. Europe still has time to change course and put telecoms at the center of its agenda — not as a technical afterthought, but as a core pillar of its defense strategy. The time for incremental steps has passed. Europe must choose to build the network foundations of its security now or accept that its strategic ambitions will remain permanently out of reach. -------------------------------------------------------------------------------- Disclaimer POLITICAL ADVERTISEMENT * The sponsor is Connect Europe AISBL * The ultimate controlling entity is Connect Europe AISBL * The political advertisement is linked to advocacy on EU digital, telecom and industrial policy, including initiatives such as the Digital Networks Act, Digital Omnibus, and connectivity, cybersecurity, and defence frameworks aimed at strengthening Europe’s digital competitiveness. More information here.

A major five-year effort to build a technology base for Europe free of U.S. influence foundered amid conflicting national strategies and powerful corporate lobbying. As Europe’s leaders once again discuss tackling American tech dependence, those involved in the project to build a European cloud warn against repeating past mistakes. The Gaia-X initiative was “a crushing failure, a colossal waste of time, and just as many years gained for the hyperscalers — in other words, an industrial disaster,” said Yann Lechelle, a former CEO of French cloud champion Scaleway and one of the founding members of the initiative who quit in frustration in 2021, describing it as the “best decision ever.” The industry-led project was born in 2019 from a Franco-German drive to forge a “European industrial policy fit for the 21st Century” — a rallying cry that brought German and French companies together with top political backing to create a data infrastructure. The endgame goal of Gaia-X, named after the Greek goddess of Earth, was to “establish data sovereignty in Europe” and “counteract monopolistic tendencies.” As political momentum once again swings behind digital sovereignty, leaders will gather in Berlin on Tuesday to talk about how to become less dependent on foreign-owned technology. POLITICO spoke to both current and former Gaia-X officials, both on and off the record, about the lessons they learned that could prove valuable. Those conversations illuminated an initiative that failed to help Europe’s own digital ecosystem take root because it was weighed down by politics, bureaucracy and the interference of precisely the American and Chinese tech titans it was meant to challenge. Despite a fast-growing market for cloud computing services that underpin the internet, the global share of European cloud providers has continued to fall, dwarfed by the dominance of Amazon, Microsoft and Google. One of Gaia-X’s initial success stories, called Agdatahub, which was touted as a triumph for farming data, went bankrupt last year. “I joined Gaia-X because I believed in the original mission. I left Gaia-X because I didn’t believe it was going in the original direction,” said its former CEO, Francesco Bonfiglio. FRANCO-GERMAN DIVIDES Misalignment among the founding companies on the mission of Gaia-X became apparent early on, consistent with the traditional divergence in Paris and Berlin over tech sovereignty. In Paris, sovereignty was about backing local champions and breaking reliance on the U.S., while Berlin focused on protecting Europe without severing important trade ties. “The influence of political happenings inside the association was evident. Sometimes they were clashing,” said Bonfiglio, describing how it pitted a “historically more protectionist” France against a “fluctuating” Germany. American cloud giants Amazon, Microsoft and Google, as well as Chinese tech giants Huawei and Alibaba, are all members of Gaia-X. | Jonas Roosens/Getty Images Everybody “interpreted” Gaia-X as they wanted to, he said. The former CEO described how this divergence in expectations and a lack of a “clear or common” definition of sovereignty — let alone a shared understanding of what it would take to get there — made his task extremely difficult. “France turned it into a very political issue, whereas the Germans treated it more as a technical matter,” said another founding member of Gaia-X, who is still part of the initiative and was granted anonymity to speak candidly. The interests were at odds from day one, founding member Lechelle recalled, which was part of the reason the initiative would never deliver “the fantasy of a European cloud Airbus.” The Germans came on board with the idea to create data sovereignty, by shielding the data of their citizens and industries from foreign snooping and legal control, he said, adding: “Atlanticist as they may be, they were totally fine with the idea of depending on Microsoft.” Meanwhile, the French pushed a more self-serving vision, hoping to see Europe become self-reliant, from infrastructure all the way to software. That’s how the mission to create a “federated cloud infrastructure” came to life. But that “staggering complexity” would soon turn into an “unmanageable mess,” said Lechelle. Current CEO Ulrich Ahle, who joined in 2023, pushed back — saying Gaia-X is far from a “failure.” It has united the industry — both large and small players — around tangible deliverables, such as federated data spaces and compliance labels, he said. “At the beginning, some people thought that Gaia-X would be the European hyperscaler as the competition to Amazon, Google, Microsoft, Alibaba and so on,” he said, but in fact, “it is more about creating a way to handle data in a European way.” “The results we’re providing and the real business benefits these interoperable data spaces are creating are more and more visible,” he said, highlighting the example of a data space based on Gaia-X standards that French energy company EDF will use to securely coordinate the construction of new nuclear sites. BACK-DOOR LOBBYING As Gaia-X grew and set out to define Europe’s blueprint for secure data sharing, it opened its doors to industry participants from beyond Europe in a bid to push new standards on the global stage. While board seats remained reserved for EU companies and industry groups, alarm bells grew louder that the project was being hijacked by the very players it was meant to take on. Those firms “steered the entire roadmap,” Lechelle said, throwing money and people at it. “The committees were drowning. They [global players] had the capacity, the bandwidth, but we were already underwater … Americans have full-time lobbyists and massive budgets. Their job is basically to derail any initiative they don’t like.” American cloud giants Amazon, Microsoft and Google, as well as Chinese tech giants Huawei and Alibaba, are all members of Gaia-X. In 2021, the annual summit in Milan was sponsored by Huawei and Alibaba, prompting backlash. Some interviewees expressed criticism that the European industry associations and companies on the board were representing the interests of business partners abroad. “I was struggling against many, many forces that were trying to dilute the rules of verification, dilute the efforts,” said Bonfiglio, stressing he was “the CEO of a consensus-based organization where consensus couldn’t be achieved most of the time.” Bonfiglio said he didn’t regret opening up the initiative to foreign players. “The problem is not America vs. Europe,” he said, but “trust” or lack thereof. Letting non-EU providers in was supposed to force them to become more transparent, he argued. “You think you’re good, show us what you have,” was his mantra at the time, he said. He now acknowledges the unavoidable influence of corporate giants in the cloud space. “You don’t need Microsoft, Amazon and Google on the board, because they would be represented by people sitting on the board from European companies. It’s an indirect lobby,” he said. The current member of the association interviewed for this story said the bylaws of Gaia-X should be changed to kick out industry associations from the board, as they play into the hands of tech giants. In response, Gaia-X’s Ahle said that “the strategic directions are given and the strategic decisions are taken in the board of directors.” He touted the initiative’s top-tier certification label — which excludes non-EU companies — as proof that it took decisions that went against U.S. interests. This was something “members like Amazon, Google and Microsoft didn’t like at all,” yet it happened. WHERE NOW As leaders prepare to meet at the high-profile summit in Berlin to debate how far to go in pivoting away from Big Tech, several of the people interviewed for this piece cautioned against repeating past mistakes. While European countries have not yet aligned on a common definition of digital sovereignty — something many see as crucial for real progress — there are signs that Paris and Germany are closer on positioning than they were five years ago. “I admit, I struggled with the term [digital sovereignty] before. I didn’t think it was necessary, but the global situation has changed so dramatically that we Europeans now have to become more sovereign,” German Chancellor Friedrich Merz said Thursday. At the summit, Merz said, “We’ll explore all the possibilities, together with industry representatives, of what we can do not only to become more independent from China, but also, for example, less dependent on the U.S., less dependent on the Big Tech companies. We want to catch up, we want to improve.” Friedrich Merz said, “We’ll explore all the possibilities, together with industry representatives, of what we can do not only to become more independent from China, but also, for example, less dependent on the U.S.” | Harald Tittel/Getty Images And yet — with Germany this month celebrating Google’s decision to invest more than €5 billion in building data centers in the country, a move that Finance Minister Lars Klingbeil described as “exactly what we need right now” — the reality of corporate interests may be hard to address. For Bonfiglio, the lesson from Gaia-X is that ”it is obvious that everybody sitting in the boardroom of an association with such a big and impactful objective tries to protect the interests of their own company.” While Gaia-X may have missed its shot at delivering on its big, original ambitions, Lechelle insists the upcoming Franco-German summit is “a chance to put a finger on the sore spots.” In the meantime, “those who wanted to maintain the status quo have won.”

BELÉM, Brazil — United Nations climate summits have for years ended with bold promises to stave off global warming. But those commitments often fade when nations go home. Three years ago, in a resort city on the Red Sea, delegates from nearly 200 countries approved what they hailed as a historic fund to help poorer nations pay for climate damages — but it’s at risk of running dry. A year later, negotiations a few miles from Dubai’s gleaming waterfront achieved the first-ever worldwide pledge to turn away from fossil fuels — but production of oil and natural gas is still rising, a trend championed by the new administration in Washington. That legacy is casting a shadow over this year’s conference near the mouth of the Amazon River, which the host, Brazil, has dubbed a summit of truth. Days after the gathering started last week, nations were still sorting out what to do with contentious issues that have typically held up the annual negotiations. As the talks opened, Brazilian President Luiz Inácio Lula da Silva said the world must “fight” efforts to deny the reality of climate change — decades after scientists concluded that people are making the Earth hotter. That led one official to offer a grim assessment of global efforts to tackle climate change, 10 years after an earlier summit produced the sweeping Paris Agreement. “We have miserably failed to accomplish the objective of this convention, which is the stabilization of greenhouse gases in the atmosphere,” said Juan Carlos Monterrey Gómez, Panama’s climate envoy and lead negotiator, during an interview at the conference site in Belém, Brazil. “Additional promises mean nothing if you didn’t achieve or fulfill your previous promises,” he added. It hasn’t helped that the U.S. is skipping the summit for the first time, or that President Donald Trump dismisses climate change as a hoax and urged the world to abandon efforts to fix it. But Trump isn’t the only reason for stalled action. Economic uncertainty, infighting and political backsliding have stymied green measures in both North America and Europe. In other parts of the world, countries are embracing the economic opportunities that the green transition offers. Many officials in Belém point to signs that progress is underway, including the rapid growth of renewables and electric vehicles and a broader understanding of both the world’s challenges and the means to address them. “Now we talk about solar panels, electric cars, regenerative agriculture, stopping deforestation, as if we have always talked about those things,” said Ana Toni, the summit’s executive director. “Just in one decade, the topic changed totally. But we still need to speed up the process.” Still, analysts say it’s become inevitable that the world’s warming will exceed 1.5 degrees Celsius since the dawn of the industrial era, breaching the target at the heart of the Paris Agreement. With that in mind, countries are huddling at this month’s summit, known as COP30, with the hope of finding greater alignment on how to slow rising temperatures. But how credible would any promises reached in Brazil be? Here are five pledges achieved at past climate summits — and where they stand now: MOVING AWAY FROM FOSSIL FUELS The historic 2023 agreement to “transition away” from fossil fuels, made at the COP28 talks in Dubai, was the first time that nearly 200 countries agreed to wind down their use of oil, natural gas and coal. Though nonbinding, that commitment was even more striking because the talks were overseen by the chief executive of the United Arab Emirates’ state-owned oil company. Just two years later, fossil fuel consumption is on the rise, despite rapid growth of wind and solar, and many of the world’s largest oil and gas producers plan to drill even more. The United States — the world’s biggest economy, top oil and gas producer and second-largest climate polluter — is pursuing a fossil fuel renaissance while forsaking plans to shift toward renewables. The president of the Dubai summit, Sultan al-Jaber, said at a recent energy conference that while wind and solar would expand, so too would oil and gas, in part to meet soaring demand for data centers. Liquefied natural gas would grow 65 percent by 2050, and oil will continue to be used as a feedstock for plastic, he said. “The exponential growth of AI is also creating a power surge that no one anticipated 18 months ago,” he said in a press release from the Abu Dhabi National Oil Co., where he remains managing director and group CEO. The developed world is continuing to move in the wrong direction on fossil fuels, climate activists say. “We know that the world’s richest countries are continuing to invest in oil and gas development,” said Bill Hare, a climate scientist who founded Climate Analytics, a policy group. “This simply should not be happening.” The Paris-based International Energy Agency said last week that oil and gas demand could grow for decades to come. That statement marked a reversal from the group’s previous forecast that oil use would peak in 2030 as clean energy takes hold. Trump’s policies are one reason for the pivot. Still, renewables such as wind and solar power are soaring in many countries, leading analysts to believe that nations will continue to shift away from fossil fuels. How quickly that will happen is unknown. “The transition is underway but not yet at the pace or scale required,” said a U.N. report on global climate action released last week. It pointed to large gaps in efforts to reduce fossil fuel subsidies and abate methane pollution. Lula opened this year’s climate conference by calling for a “road map” to cut fossil fuels globally. It has earned support from countries such as Colombia, Germany, Kenya and the United Kingdom. But it’s not part of the official agenda at these talks, and many poorer countries say what they really need is funding and support to make the shift. TRIPLE RENEWABLE ENERGY, DOUBLE ENERGY EFFICIENCY This call also emerged from the 2023 summit, and was considered a tangible measure of countries’ progress toward achieving the Paris Agreement’s temperature targets. Countries are on track to meet the pledge to triple their renewable energy capacity by 2030, thanks largely to a record surge in solar power, according to energy think tank Ember. It estimates that the world is set to add around 793 gigawatts of new renewable capacity in 2025, up from 717 gigawatts in 2024, driven mainly by China. “If this pace continues, annual additions now only need to grow by around 12 percent a year from 2026 to 2030 to reach tripling, compared with 21 percent originally needed,” said Dave Jones, Ember’s chief analyst. “But governments will need to strengthen commitments to lock this in.” The pledge to double the world’s energy efficiency by 2030, by contrast, is a long way behind. While efficiency improvements would need to grow by 4 percent a year to reach that target, they hit only 1 percent in 2024. ‘LOSS AND DAMAGE’ FUND When the landmark fund for victims of climate disasters was established at the 2022 talks in Sharm El-Sheikh, Egypt, it offered promise that billions of dollars would someday flow to nations slammed by hurricanes, droughts or rising seas. Three years later, it has less than $800 million — only a little more than it had in 2023. Mia Mottley, prime minister of Barbados, excoriated leaders this month for not providing more. Her rebuke came little more than a week after Hurricane Melissa, one of the strongest tropical cyclones ever seen in the Atlantic, swept across the Caribbean. “All of us should hold our heads down in shame, because having established this fund a few years ago in Sharm El-Sheikh, its capital base is still under $800 million while Jamaica reels from damage in excess of $7 billion, not to mention Cuba or the Bahamas,” she said. Last week, the fund announced it was allocating $250 million for financial requests to help less-wealthy nations grapple with “damage from slow onset and extreme climate-induced events.” The fund’s executive director, Ibrahima Cheikh Diong, said the call for contributions was significant but also a reminder that the fund needs much more money. Richard Muyungi, chair for the African Group of Negotiators and Tanzania’s climate envoy, said he expects additional funds will come from this summit, though not the billions needed. “There is a chance that the fund will run out of money by next year, year after next, before it even is given a chance to replenish itself,” said Michai Robertson, a senior finance adviser for the Alliance of Small Island States. GLOBAL METHANE PLEDGE Backed by the U.S. and European Union, this pledge to cut global methane emissions 30 percent by 2030 was launched four years ago at COP26 in Glasgow, Scotland, sparking a wave of talk about the benefits of cutting methane, a greenhouse gas with a relatively short shelf life but much greater warming potential than carbon dioxide. “The Global Methane Pledge has been instrumental in catalyzing attention to the issue of methane, because it has moved from a niche issue to one of the critical elements of the climate planning discussions,” said Giulia Ferrini, head of the U.N. Environment Program’s International Methane Emissions Observatory. “All the tools are there,” she added. “It’s just a question of political will.” Methane emissions from the oil and gas sector remain stubbornly high, despite the economic benefits of bringing them down, according to the IEA. The group’s latest methane tracker shows that energy-based methane pollution was around 120 million tons in 2024, roughly the same as a year earlier. Despite more than 150 nations joining the Global Methane Pledge, few countries or companies have devised plans to meet their commitments, “and even fewer have demonstrated verifiable emissions reductions,” the IEA said. The European Union’s methane regulation requires all oil and gas operators to measure, report and verify their emissions, including importers. And countries and companies are becoming more diligent about complying with an international satellite program that notifies companies and countries of methane leaks so they can repair them. Responses went from just 1 percent of alerts last year to 12 percent so far in 2025. More work is needed to achieve the 2030 goal, the U.N. says. Meanwhile, U.S. officials have pressured the EU to rethink its methane curbs. Barbados and several other countries are calling for a binding methane pact similar to the Montreal Protocol, the 1987 agreement that’s widely credited with saving the ozone layer by phasing out the use of harmful pollutants. That’s something Paris Agreement architect Laurence Tubiana hopes could happen. “I’m just in favor of tackling this very seriously, because the pledge doesn’t work [well] enough,” she said. CLIMATE FINANCE In 2009, wealthy countries agreed to provide $100 billion annually until 2025 to help poorer nations deal with rising temperatures. At last year’s climate talks in Azerbaijan, they upped the ante to $300 billion per year by 2035. But those countries delivered the $100 billion two years late, and many nations viewed the new $300 billion commitment with disappointment. India, which expressed particular ire about last year’s outcome, is pushing for new discussions in Brazil to get that money flowing. “Finance really is at the core of everything that we do,” Ali Mohamed, Kenya’s climate envoy, told POLITICO’s E&E News. But he also recognizes that governments alone are not the answer. “We cannot say finance must only come from the public sector.” Last year’s pledge included a call for companies and multilateral development banks to contribute a sum exceeding $1 trillion by 2035, but much of that would be juiced by donor nations — and more countries would need to contribute. That is more important now, said Jake Werksman, the EU’s lead negotiator. “As you know, one of the larger contributors to this process, the U.S., has essentially shut down all development flows from the U.S. budget, and no other party, including the EU, can make up for that gap,” he said during a press conference. Zack Colman and Zia Weise contributed to this report from Belém, Brazil.

NEWPORT, Wales — Road signs around Newport still refer to this sprawling former industrial site as a radiator factory. But soon, it will generate a different kind of heat. Microsoft has chosen this area of South Wales — once the world’s steel capital — to build hulking new data centers. Five buildings, covering an area larger than three football pitches, are springing up to meet what the company describes as “exploding demand” for artificial intelligence compute power.  For Microsoft, the area’s industrial heritage is precisely why it’s investing. Newport’s legacy of heavy-duty factories means it has the infrastructure needed for energy-intensive data centers. But doubts over whether Britain can supply enough energy to keep up with demand from data centers are an urgent problem for the government’s AI ambitions. The government’s former AI adviser Matt Clifford has warned that without energy and planning reform, new data center projects and the billions of pounds of investment they bring are at risk.  Britain’s industrial electricity prices are 60 percent higher than the average of countries in the International Energy Agency, and waits for a grid connection can stretch to a decade.  “We had the biggest AI funders in the world lining up to invest tens of billions into our infrastructure if only we could sort out our energy mess,” Clifford said at an event about his time in No.10. U.S. Ambassador to the U.K. Warren Stephens, Donald Trump’s point man in London, is also watching closely, calling Britain’s energy costs the country’s “chief obstacle” to growth. “If there are not major reforms to U.K. energy policy, then the U.K.’s position as a premier destination in the global economy is vulnerable,” Stephens warned a business gathering in London. A TALL ORDER The Newport project will need 80MW of energy – enough to power a small town – but the Department for Science, Innovation and Technology (DSIT) predicts the country needs to boost its total data center capacity five-fold by 2035, from 1.8GW to 9.6GW. That expansion will mean data centers’ power total demand will treble over the same period, according to NESO, the body which manages U.K. electricity demand. A spokesperson for the DSIT said it was looking at “bespoke options” to support data centers’ energy demands, adding: “The work of our AI Energy Council — bringing together regulators, energy companies and tech firms — will ensure we can do that using responsible, sustainable sources.” AI Minister Kanishka Narayan told a conference for AI researchers in London in October that there was “no better place to build” than Britain, arguing its combination of talent, access to capital and large public markets is unmatched. Investors aren’t so sure.  “People aren’t willing to pay a premium on U.K. power rates to run their workloads here,” Mike Mattacola, international general manager at AI infrastructure company CoreWeave said at the same conference. “We need to fix that.” SELLING THE SHOVELS It’s not just energy prices that are the problem. The boss of Hitachi Energy U.K., which is working with the National Grid to upgrade Britain’s power network, warned that the grid is the biggest hurdle to Britain’s AI ambitions. Laura Fleming said data centers should be at “the heart” of the country’s energy planning, but added: “I’m still not sure whether as the U.K. we have sufficiently planned for this.”  More than half all applications for a grid connection are now made by data centers, according to the National Grid. Energy regulator Ofgem is trying to get a grip of things, grumbling that amid the “credible data center projects” applying for a grid connection, they want to get rid of “less viable projects that may crowd out those with genuine merit.” Power providers, meantime, are lining up to find the opportunities in this uncertainty. Two hundred miles to the north of Newport, the U.K.’s largest power station is offering itself as one solution. Drax Power Station burns wood pellets imported from North America and wants to build data centers hooked up to its four biomass terminals.  Richard Gwilliam, director of future operations, revealed that Drax has already held talks with hyperscalers and plans to bring a data center online in the early 2030s. He hoped the 2.6-gigawatt power station could offer “big scale stuff” to the market. Gwilliam also said the existing connections gave biomass a trump card to play in the data center race. SQUARING THE CIRCLE The rush for power is also clashing with Britain’s net zero ambitions. The most in-demand energy source for data centers is still fossil fuels, specifically gas. National Gas said it has had inquiries from five big data center projects since last November, equivalent to 2.5GW worth of energy capacity, or twice the capacity of Britain’s biggest nuclear power station, Sizewell B.  Its chief commercial officer, Ian Radley, argued gas provided customers with “the flexibility and capacity they need to enable the Government’s strategic AI ambitions.”  But environmental groups point out that the surge in carbon emissions from new data centers have not been factored in to the  U.K.’s Carbon Budget Delivery Plan, which sets out a path for the government to hit legally-binding climate goals up to 2037.  “It’s unclear how the government intends to square the circle of encouraging a construction frenzy of new, highly polluting data centers while not overshooting the binding climate targets they need to meet,” said Donald Campbell, director of advocacy at campaign group Foxglove. This tension is also being played out at the AI Energy Council, a body the government formed in January to bring AI and energy companies together, but which has only met twice.  It is co-chaired by two ministers with different priorities. Ed Miliband, as energy secretary, needs to cut Britain’s emissions to zero by 2050, while Technology Secretary Liz Kendall needs to turn AI’s promises of investment and growth, particularly to left-behind areas, into a reality.  The government has pushed the idea AI Growth Zones — huge data center campuses on former industrial land, which already have grid connections and will get fast-tracked through planning — as a solution. One has already been announced in Northumberland, but a decision on a second, planned for Teesside in north-east England, has been delayed until the end of this year by Miliband, whose department has to make a call on whether to greenlight plans for a hydrogen plant on the same site, which could preclude data centers being built there. “There is a large fight going on inside of government where Ed Miliband seems to have set himself up against not just the prime minister, but a number of secretaries of state,” Houchen told POLITICO during Conservative Party Conference in October.  THE NUCLEAR OPTION Long term, the government is betting on a cleaner, but more expensive energy source — nuclear, specifically small modular reactors. Michael Jenner, CEO of nuclear firm Last Energy UK, said they had received dozens of enquiries from data center builders and argued that the green credentials of nuclear was an ace card it could play against rival bids from gas companies.   “If you’re thinking about building data centers in South Wales, which a lot of people are, you have a problem with the authorities because they don’t want new gas there,” he said.  In September, EDF Energy announced plans to work with American company Holtec International building a crop of data centers next to small modular nuclear reactors at a disused coal plant in Nottinghamshire.  The Tony Blair Institute, which is influential with government ministers, has argued nuclear has a “unique” advantage when it comes to data centers. It also believes the country should scale back its net zero plans in favor of reducing energy costs to attract data center investment.  “Cheap, firm power is … not a ‘nice to have’ but a prerequisite for attracting AI-driven growth,” it argued in a report last month. Gas, meanwhile, should be part of that energy mix, the Institute recommended in July. Firms represented at the AI Energy Council have urged ministers to green-light greater use of gas turbines in the short term. The clock is ticking. Gas, nuclear, renewables or even wooden pellets — ministers willing on an AI revolution need to make decisions fast.  

LONDON — The U.K. government is going all-out to get Brits putting their money in stocks and shares. The timing could definitely be better. Lead policymakers and City of London analysts are increasingly warning of an artificial intelligence-fueled correction in equities just as the U.K.’s top finance minister prepares a major new policy to push Britain’s savers into the stock market. Chancellor Rachel Reeves has made upping retail participation in stocks and shares a high priority, launching a campaign earlier this year to unite financial firms in an advertising blitz extolling the benefits of investing. At next month’s budget, she’s expected to push changes to the tax system that would encourage investors to swap their steady, tax-free cash savings products for a stocks and shares ISA. With AI stocks soaring, it’s caused some raised eyebrows in the City. AI stocks in the U.S. account for roughly 44 percent of the S&P 500 market capitalization, and Nvidia just became the first company in history to become worth $5 trillion. The meteoric rise in has led some experts to warn there’s only one way out: The bubble will burst. “It would, unfortunately, be poetic timing if a major correction arrives just as the government is trying to get more people into investing,” said Chris Beauchamp, chief market analyst at IG. ATLANTIC INFLUENCE This week, City broker Panmure Liberum found that 38 percent of the U.S. stock market’s value is based in a “speculative component” that AI companies will continue to build out data centers and spend billions more on chips — by no means a sure bet. “While this capital spending could deliver substantial productivity gains that might eventually spread to the broader market, there is still no clear evidence that this is happening and is difficult to forecast the size of an eventual impact,” said Panmure analyst Susana Cruz in a research note. The “Magnificent Seven” group of tech giant composed around 20 percent of the S&P 500 at the end of 2022, but now make up more than a third of it, having tripled in size over just three years. The American index’s price-to-book ratio (meaning a company’s market cap compared to assets and liabilities) is at an all-time high, with 19 of the 20 valuation metrics tracked by Bank of America more expensive than the historical average. Despite the vast valuations, an infamous MIT study published earlier this year found that 95 percent of companies using generative AI were getting zero return. In early October, the Bank of England’s committee which monitors risks to financial stability warned of a “sudden correction” in markets, saying that “equity valuations appear stretched” as valuation metrics reached levels comparable to the peak of the dotcom bubble that unfolded in the early millennium, when the Nasdaq fell 77 percent from its peak, wiping trillions of the stock market. It took 15 years for the index to recover. The U.K. central bank’s warning came a month after global body, the Bank for International Settlements, issued a similar caution. Kristalina Georgieva, head of the International Monetary Fund, has also drawn comparisons with the dotcom bubble. Even Jamie Dimon, chief executive of U.S. banking giant JP Morgan, has said he’s seriously worried about a market correction. Over most periods investment beats cash, as long as individuals are willing to lock their money away for several years. Savers could have doubled their money over the last decade by putting their cash in the stock market rather than keeping it in a savings account, according to Schroders. Nvidia is up 13 percent this month alone — rather than an index fund which tracks hundreds of stocks, they stand to lose a lot of money if things go sour. | Jung Yeon-Je/Getty Images “No one can time the market, definitely not a bulky institution like the government,” Oliver Tipping, analyst at investment bank Peel Hunt, said. “Big picture, the government is right to try to stimulate more retail investment.” But if an individual decides to put their hard-earned savings into stocks they perceive as doing particularly well — Nvidia, for example, is up 13 percent this month alone — rather than an index fund which tracks hundreds of stocks, they stand to lose a lot of money if things go sour. “If you think about your average Joe, they’re not going to go into a safe index fund, they’ll put all of their money in Nvidia or Facebook and could get in at the wrong time,” one financial analyst, granted anonymity to speak freely, said.  Yet even an index fund, like a global equities tracker, is made up of close to 20 percent of the “Magnificent Seven” companies, due to the massive size of the American stock market compared to the rest of the world. While these funds have suffered significant drops in the past — U.S. President Donald Trump’s threat of tariffs in April caused a drop of 10 percent in a week — they have then recovered over a period of months or years. That’s good news for investors willing to wait for the market to correct any possible downturn — but if retail investors panic and withdraw their funds at the first sign of a loss, they could end up with less money than they put in, possibly wiping out emergency savings. “There is clearly a risk here that government is pushing people to invest when maybe they don’t have enough of a cash buffer in order to do that, that you’re going to be setting up problems for the long term, and it’ll be interesting to see who’s on the hook for paying that compensation,” said Debbie Enver, head of external affairs at the Building Societies Association. ONCE BITTEN, TWICE SHY City analysts also express concern that investors entering the stock market for the first time could be forever turned off from shifting their cash over to equities if an immediate correction is nigh. Only 8 percent of wealth held by U.K. adults is in stocks and funds, four times lower than in the U.S., according to data from asset manager Aberdeen. “There is no doubt that the government would find it much harder to drive retail investment in a period of financial turbulence,” added Chris Rudden, head of investment consultants at Moneyfarm. “Appetite to invest is linked to strong recent market performance. If there was to be a bubble that bursts in the coming few months, then it could make their job impossible.” IG’s Beauchamp argued that the government would need to pursue a broader education plan “to help people through the inevitable pullback” and prevent them from avoiding the stock market permanently. “How you do that without scaring people witless is a Herculean task,” he added. Laith Khalaf, head of investment analysis at AJ Bell, suggested investment platforms could encourage regular incremental savings in the stock market, known as dollar cost averaging, rather than throwing one lump sum in, which he said “mitigates the risk of a big market downdraft.” One solution that appears to be under consideration by Reeves as part of the autumn budget is to introduce a minimum U.K. stock shareholding in ISAs — which she could argue would protect British savers from a U.S. downturn and pump more money into local companies. This too is not without risk. The FTSE 100 derives nearly 30 percent of its revenue from the U.S., according to the London Stock Exchange, and U.K. markets are generally incredibly sensitive to macroeconomic shifts across the Atlantic. The FTSE 100 derives nearly 30 percent of its revenue from the U.S., according to the London Stock Exchange. | Jeff Moore/Getty Images Meanwhile, if an AI-induced stock bubble isn’t enough cause for concern, worries of trouble in the private credit sector exploded this month after the collapse of sub-prime auto lender Tricolor and car parts supplier First Brands left some U.S. banks with significant losses, causing a spillover onto public markets. BoE governor Bailey recently drew similarities between risks in the asset class and the 2008 global financial crisis, saying it was an “open question” if the event was “a canary in the coal mine” for a market meltdown. If one domino falls, they all could — and that would leave Britain’s chancellor in a real bind.

AI is intensifying the strategic rivalry between the European Union and the United States, reshaping models of industrial policy and regulatory sovereignty. Amid a flurry of investment announcements, the exposure of security vulnerabilities and the contest over global standards, one critical factor remains largely in the shadows — seldom acknowledged, scarcely quantified and rarely debated: its environmental footprint. The environmental blind spot of a strategic technology The silence surrounding the impact of AI is surprising. A study carried out by Sopra Steria and Opsci.ai analyzing over 3 million posts about AI on social media reveals that its environmental impact accounts for less than 1 percent of the global conversation.1 Worse still, among the 100 most influential AI personalities,2 ecological concerns are only eighth on the list of subjects they discuss most, far behind technological and economic issues. > A study carried out by Sopra Steria and Opsci.ai analyzing over 3 million > posts about AI on social media reveals that its environmental impact accounts > for less than 1 percent of the global conversation AI relies on energy-intensive infrastructure that consumes resources and water, the footprint of which remains largely underestimated, poorly measured and therefore little considered in industrial and political trade-offs. This misalignment can also be explained by the trajectory of the sector itself: driven by the rise of AI, the digital sector is one of the few areas whose environmental impact is continuing to grow, contrary to the climate objectives set out in the Paris Agreement. While American players are already crushing the AI market, technological dependence must not be compounded by a setback on Europe’s carbon trajectory. This omission undermines the credibility of any European industrial strategy built on AI. To serve as genuine drivers of transformation, the leading AI companies must bring full transparency to their environmental trajectory — one they are progressively shaping for Europe. © Sopra Steria Measuring for action: The need for transparency and rigor We must not rush to condemn AI, but we must insist on setting the conditions for its long-term sustainability. This means measuring its impact objectively and transparently, equipping stakeholders with the tools for informed debate, and guiding decision-makers in their technological choices. Recent research indicates that the environmental footprint of a given model can vary significantly depending on where it is assessed, the energy mix of the countries hosting the data centers,3 the duration of the training, the architecture employed and the extent to which low-carbon energy sources are used. Breaking through the methodological vagueness means providing developers, purchasers and decision-makers with common frames of reference, impact simulators, libraries of low-carbon models and low-carbon computing infrastructures. Numerous levers for action and choice exist, provided we have the necessary data and tools. This requirement is not a regulatory whim but a strategic steering tool. Sustainability must be given as much weight as performance or security in industrial and economic trade-offs, because it determines the very viability of Europe’s strategic autonomy. At a time when free international trade faces headwinds, and as the second phase of the AI Act — in force since August 2025 — continues to overlook environmental sustainability, transparency on environmental impact must become a prerequisite for access to European markets, financing and large-scale deployment. Making sustainability a central pillar of European competitiveness Europe has an opportunity to seize. It has a robust standards base that is a powerful lever for competitiveness and responsible innovation, provided that it is supported by targeted investment, shared standards and an industrial strategy aligned with our climate objectives. But Europe can rely on something even more decisive: its people. We have world-class researchers, visionary entrepreneurs, and thriving companies that embody the best of technological and industrial excellence. The recent strategic partnership between ASML, a key supplier to the world’s semiconductor industry, and Mistral, an AI start-up, illustrates Europe’s capacity to connect its industrial and digital strengths to shape a sovereign and sustainable future4. It would be dangerous to suggest that Europe’s technological strength could be built on deferred ecology. What is tolerated as a gray area today will be a competitive handicap tomorrow. Customers, investors and citizens will increasingly demand transparency. The emergence of responsible AI does not mean making it perfect, but making it readable, controllable and adjustable. In a technological landscape dominated by two superpowers that have hitherto favored efficiency and technological competitiveness to the detriment of ethical safeguards, Europe can chart a singular course. It has the means to assert itself by defending responsible AI, at the service of the common good and in line with its fundamental values: the rule of law, individual freedom, social justice and respect for the environment. This orientation is not a brake on innovation, but on the contrary a lever for differentiation, capable of inspiring confidence in a digital ecosystem that is often perceived as opaque or threatening. By betting on ethical, explainable and sustainable AI, Europe would not be giving up global competition, but it would be redefining the rules of the game. More than ever, it must give priority to clarity, stringency and rigor. Only then will AI cease to be a technological equation to be solved and become a genuine project at the service of our society, consistent with our democratic and ecological imperatives. -------------------------------------------------------------------------------- 1. AI & environment: breaking through the information fog – Sopra Steria 2. “The 100 Most Influential People in AI 2024”, Time Magazine 3. ADEME – Arcep study on the environmental footprint of digital technology in 2020, 2030 and 2025 4. https://www.politico.eu/article/dutch-asml-invests-in-french-mistral-in-huge-european-ai-team-up/