The European Union is moving closer to a binding crackdown on Chinese-made technology in sensitive sectors, signaling a sharp escalation in its effort to reduce strategic dependence on Beijing amid rising geopolitical tensions and intensifying pressure from Washington.

According to the Financial Times, Brussels is preparing a proposal that would force EU member states to phase out equipment from companies deemed “high-risk vendors,” including Huawei and ZTE, from critical infrastructure such as telecommunications networks and solar energy systems. The plan would effectively end the bloc’s current voluntary approach and replace it with a mandatory regime under its broader cybersecurity framework.

If adopted, the shift would mark one of the EU’s most consequential security-driven technology interventions to date, with far-reaching implications for telecom operators, energy providers, supply chains, and EU–China relations.

From voluntary guidelines to binding rules

Since 2020, the EU has relied on a non-binding “toolbox” that encourages member states to assess and mitigate risks from foreign vendors in 5G and other critical networks. While some countries, including Sweden and Denmark, moved quickly to restrict Chinese suppliers, others took a far more cautious stance.

Large economies such as Germany and Spain resisted outright bans, citing the cost of replacing existing equipment, potential disruptions to network stability, and the lack of sufficient alternatives at scale. As a result, Huawei equipment remains deeply embedded in parts of Europe’s telecom infrastructure, particularly in legacy 4G networks that underpin 5G roll-outs.

EU officials now appear convinced that the voluntary model has reached its limits. The proposed cybersecurity initiative, expected to be unveiled on Tuesday, would compel governments to act, closing loopholes that have allowed national discretion to override bloc-wide security concerns.

Notably, the scope of the proposal goes beyond telecommunications. By extending restrictions to solar energy systems, the EU is signaling that it now views energy infrastructure through the same national security lens as digital networks.

China dominates large segments of the global solar supply chain, from polysilicon to finished panels, and European policymakers have grown increasingly uneasy about that reliance as renewable energy becomes central to economic and industrial strategy. For Brussels, the issue is no longer just cybersecurity, but long-term resilience, control over critical inputs, and protection against coercive leverage.

The timelines for phasing out Chinese equipment would vary depending on the sector, perceived risk, and the availability of alternative suppliers. Officials cited in the report said cost considerations would also be factored in, suggesting a gradual transition rather than an abrupt dismantling of existing infrastructure.

Industry resistance and economic trade-offs

Telecom operators are likely to push back. Industry groups have repeatedly warned that forced equipment replacement could cost billions of euros, slow network upgrades, and weaken Europe’s competitiveness in 5G and future 6G technologies. Many operators are already grappling with heavy capital expenditure, weak revenue growth, and rising energy costs.

Replacing Chinese equipment is not simply a matter of switching vendors. Networks are deeply integrated systems, and removing one supplier often requires broader redesigns. European alternatives such as Ericsson and Nokia stand to benefit commercially, but scaling up quickly enough to meet demand remains a challenge.

There are also concerns that higher infrastructure costs could ultimately be passed on to consumers or delay digital inclusion goals, particularly in rural and underserved regions.

Alignment with the United States

The EU’s move brings it closer to the U.S. position, which has taken a far more aggressive stance. Washington banned approvals of new telecommunications equipment from Huawei and ZTE in 2022 and has consistently urged allies to exclude Chinese vendors, framing the issue as a matter of national security and intelligence protection.

While European officials have often bristled at what they see as American pressure, the convergence of policy suggests a growing alignment driven by shared threat assessments rather than diplomacy alone.

Strained EU–China relations

For Beijing, the proposal is likely to be viewed as another sign that Europe is hardening its stance. China has previously accused the EU of politicizing trade and discriminating against Chinese companies, warning that such measures could damage economic ties.

Huawei, once a symbol of deep technological cooperation between China and Europe, has already felt the chill. The company has been reassessing the future of a newly completed manufacturing plant in eastern France, amid regulatory uncertainty, slow 5G deployment, and shrinking market access across the continent.

Taken together, the proposal reflects a broader recalibration of EU policy toward “economic security.” Brussels has increasingly emphasized de-risking rather than full decoupling, but mandatory exclusions from critical infrastructure represent a more forceful use of regulatory power.

The challenge now lies in execution. Member states must balance security priorities with economic realities, manage industry fallout, and ensure that Europe does not simply swap one dependency for another.

The artificial intelligence boom is no longer just a story of rapid breakthroughs and bold promises. It is increasingly a story about money — staggering amounts of it — and how long investors are willing to keep writing cheques before profits appear.

Across the industry, companies are spending tens of billions of dollars on ever-larger models, sprawling data centers, and specialized chips, all in pursuit of dominance in what is widely described as a once-in-a-generation technological shift.

For now, the narrative still holds. AI, its backers argue, will eventually transform productivity, business models, and entire economies. That belief has been strong enough to support eye-watering valuations and justify losses that would be unthinkable in most other sectors. But as spending accelerates, a harder question is pushing its way to the surface: who can afford to stay in the race long enough to win?

That question hangs most heavily over OpenAI, the company that did more than any other to bring generative AI into the mainstream. Since the release of ChatGPT just over three years ago, OpenAI has become a household name and a central force in the global AI conversation. Yet behind the visibility and influence lies a financial position that many analysts now see as increasingly precarious.

Unlike rivals such as Google and Meta, OpenAI does not have a mature, cash-generating core business to fund its ambitions. Google can lean on advertising and cloud computing. Meta can tap profits from its social media empire. Both can afford to pour hundreds of billions of dollars into AI over many years, even if returns are slow to materialize. OpenAI cannot. It survives on external funding, partnerships, and the hope that scale will eventually unlock a sustainable business.

That has not stopped the company from committing to extraordinary levels of spending. OpenAI is expected to lay out well over $1 trillion before the end of the decade, largely on computing infrastructure and model development. It is a bet that size and speed will prove decisive, even as revenue lags far behind costs.

Subscription uptake for ChatGPT has been weaker than many early forecasts suggested, highlighting users’ limited willingness to pay directly for AI tools. While the company has begun exploring enterprise services, licensing deals, and other commercial avenues, those efforts are still in their infancy.

The imbalance between spending and income has sharpened concerns about how long OpenAI can keep burning cash. In a recent essay for the New York Times, quoted by Yahoo Finance, Sebastian Mallaby, a senior fellow at the Council on Foreign Relations, warned that the company could run out of money “over the next 18 months.” His argument is rooted less in skepticism about AI itself and more in the brutal economics of the race now unfolding.

Mallaby is not dismissive of the technology. On the contrary, he is bullish, arguing that while new technologies usually take decades to be deployed effectively, AI has made “striking” progress in just three years. His analysis instead focuses on competitive advantage. Companies with deep, profitable legacy businesses can afford to treat AI as a long-term investment. OpenAI, without that cushion, must repeatedly return to capital markets to fund losses that are already enormous.

Those losses are mounting fast. Despite raising record sums for a private company, OpenAI is estimated to have burned through more than $8 billion in 2025 alone. Mallaby argues that even if the firm scales back some commitments or uses its highly valued shares to offset certain costs, it still faces a daunting funding gap. The scale of capital required, he suggests, may simply exceed what investors are willing to provide indefinitely.

If funding dries up, consolidation becomes the most likely outcome. Mallaby suggests OpenAI could be absorbed by a cash-rich technology giant such as Microsoft or Amazon, effectively ending its existence as an independent player. Such a scenario would not necessarily mark a failure of AI as a technology. Instead, it would underline how capital-intensive the industry has become and how difficult it is for standalone firms to compete with tech behemoths that can afford years of losses.

That distinction matters. Even in a collapse or takeover scenario, OpenAI’s influence would be hard to erase. The company helped set the pace of innovation, forced competitors to accelerate their own AI efforts, and reshaped public expectations of what machines can do. As Mallaby puts it, the failure of OpenAI would not be an indictment of AI, but rather the end of what he describes as the most hype-driven builder of it.

Others across the industry share the sense that a reckoning is approaching. Several analysts describe 2026 as a make-or-break year for OpenAI, as investor patience wears thin and competition intensifies. Pressure is also rising on the broader AI sector to show clearer paths to profitability, particularly as interest rates and macroeconomic uncertainty make easy money harder to come by.

Sam Altman, OpenAI’s chief executive, shows no sign of backing down. He has reportedly declared “code red” internally and is doubling down on ChatGPT, determined to keep pace with Google, which is rapidly closing the gap with its own AI models. The strategy suggests a belief that retreat would be more dangerous than pressing ahead.

However, the efforts have done so little to quell growing skepticism. One venture capital executive, who invested in a rival AI firm, recently described OpenAI’s trajectory to The Economist as “the WeWork story on steroids,” invoking a company that expanded aggressively on the back of hype and capital before collapsing under the weight of its own business model.

Against this backdrop, it is becoming clear that the AI boom is entering a more unforgiving phase, and the focus is shifting from what the technology can do to what it can earn.