US-China Trade Truce Dies — And the Tech World Won't Survive It Unchanged

The deal that wasn’t

The truce is over. In New York, US and Chinese ministerial negotiators sat down hoping to extend the fragile pause on the tariff and export-control war that has been grinding through the global economy since 2024. According to the Financial Times, citing US officials, no agreement was reached. The high tariffs stay. The export controls stay. The pause — a temporary de-escalation that gave companies breathing room to retool supply chains — expires without renewal.

What happened in that room matters far more than the press release. For the first time in two years, there is no off-ramp explicitly on the table between the world’s two largest economies. That changes how every multinational operates, from semiconductor fabs to AI research labs to mining operations in southern China.

Who wins, who loses

China’s mineral-processing advantage wins. Beijing controls roughly 60 per cent of global rare-earth mining and nearly 90 per cent of processing capacity. When the truce held, companies could plan around it. Now, the absence of an extension signals that Washington sees no reason to moderate its posture, and Beijing sees no reason to blink. The leverage stays where it always was — in the ground and the refineries of Inner Mongolia and Bayan Obo.

American chip designers and foundries lose the most in the near term. Export controls on advanced AI chips to China were the centerpiece of the truce framework. Without an extension, those restrictions harden. Companies that built their China revenue on A100 and H20 GPUs — Nvidia, AMD, Intel’s foundry clients — face an immediate ceiling. The Pentagon’s CHIPS Act subsidies don’t replace lost Beijing revenue overnight. They build fabs in Arizona and Taiwan. They take three to five years to come online. Meanwhile, Chinese firms accelerate their own alternatives, funded by state policy that now has no reason to wait.

Japan and South Korea occupy an uncomfortable middle ground. The same day the truce collapsed, Japanese, American, and South Korean foreign ministers issued a joint statement opposing “economic coercion,” according to Kyodo News. They also agreed to a new trilateral partnership on Arctic cooperation — a move that sounds symbolic but signals a broader alignment the US is pushing: decouple from China where it counts, coordinate everywhere else. For Tokyo and Seoul, that means choosing between deepening ties with Washington and losing access to the Chinese market, which accounts for roughly 20 per cent of their exports combined. The truce’s end makes that choice harder, not easier.

The second-order shock: AI governance

The export-control question isn’t just about semiconductors. It’s about who governs artificial intelligence. The truce had implicitly accepted a division of labour: the US would restrict hardware exports to China; China would restrict data flows out of its territory. Both sides traded control for control. Without an extension, that informal architecture dissolves.

The US is likely to push for tighter controls on dual-use AI research and training clusters — not just the chips themselves, but the models that run on them. China will respond by treating AI as a strategic resource the state controls directly, tightening domestic oversight and accelerating indigenous model development. The gap in compute access widens. The gap in governance approaches widens faster.

This isn’t abstract. European companies that have bet on a multilateral AI framework — the EU AI Act’s enforcement phase is underway — suddenly find themselves operating in a world where the two largest markets have abandoned even a partial consensus. The Brussels effect, which once depended on the US and China eventually converging on standards, loses its anchor.

Rare earths as the quiet threat

The most dangerous consequence is also the least discussed. The truce didn’t formally address rare-earth export restrictions. China hasn’t needed to — the mere threat has been enough to keep Western manufacturers hedging. With the truce dead, Beijing has removed the last diplomatic incentive to restrain itself. Australia, Vietnam, and Madagascar are developing alternative sources, but none will reach commercial scale before 2028. The bottleneck stays.

Defense contractors feel this first. F-35 engines, missile guidance systems, submarine sonar — all depend on refined neodymium and praseodymium. The Pentagon has been stockpiling for two years. It won’t be enough.

What happens next

Expect the US to respond with its own escalation, not retreat. Treasury is likely to expand secondary sanctions on entities that help China circumvent chip controls — a move that will hit Japanese and European distributors directly. Commerce will widen the entity list for AI training clusters. Congress may fast-track legislation that ties CHIPS Act disbursements to stricter export-enforcement benchmarks.

China will respond symmetrically. Expect additional rare-earth export quotas targeting specific countries, new restrictions on gallium and germanium — materials critical to EV batteries and radar systems — and possibly a review of US companies operating in China, from Apple to Tesla. The retaliation won’t come as a single announcement. It will come as a series of bureaucratic measures, each small enough to deny a trade war label but cumulative enough to reshape entire industries.

The 2026 G20 summit, scheduled for later this year, becomes the next probable venue for crisis management. But without a truce to extend, the baseline has shifted. The world is no longer managing a controlled conflict. It is entering an unmanaged one.

For companies that spent 2024 and 2025 building contingency plans around the assumption that the US and China would eventually sit down again, the assumption is now wrong. The truce was never meant to last. But it lasted long enough for some to forget that. It won’t last long enough for anyone to recover from its absence.

The broader signal

The parallel diplomacy on display — the US-Japan-ROK foreign-minister meeting, the Arctic partnership proposal — confirms what the trade failure implies. America is no longer trying to keep China inside the system it built. It is rebuilding the system around allies who agree to exclude China from critical nodes. Whether that strategy succeeds depends on whether those allies can afford the friction. Right now, the arithmetic is unclear.

The truce’s collapse is not the start of a new trade war. It is the end of the illusion that the last one could be paused. The real contest — over chips, models, minerals, and the rules that govern them — begins now.