Trump and Xi Meet a Third Time in Eleven Months. The Truce Between Them Still Has No Signature.

The sequence starts with escalation, not diplomacy. By spring 2025, US tariffs on Chinese goods had climbed to roughly 145% at their peak, and China's response was to tighten rather than negotiate: it moved toward broader rare-earth export controls rather than making trade concessions. The two governments met at Gimhae International Airport near Busan on 30 October 2025, and the White House's fact sheet of 1 November 2025 laid out the terms in writing: the US would lower its fentanyl-related tariff on Chinese imports from 20% to 10% effective 10 November, while the pre-existing 10% reciprocal tariff would be suspended through 10 November 2026 — one year, not indefinitely. China committed to buying 12 million metric tons of US soybeans before the end of 2025 and at least 25 million tons annually through 2028, to resuming sorghum and log purchases, and to suspending its own retaliatory tariffs on US chicken, wheat, corn, cotton, pork, beef and dairy. On rare earths specifically, China agreed to suspend new export controls and issue general licenses for rare earths, gallium, germanium, antimony and graphite — a commitment later formalized as running to 27 November 2026. Atlantic Council analysts reviewing the Busan outcome called it 'real results' that mostly restored the position both sides held before the spring's escalation, not a net gain for either. The second summit, in Beijing in mid-May 2026, was framed by MercoPress as turning on five fronts at once — Iran (Trump wanted Chinese pressure on Tehran), Taiwan (Beijing's 'core interest'), tariffs (technical talks continuing in South Korea, the US pushing Boeing and agricultural purchases), rare earths (a bare US request for continued cooperation), and AI safety (a token first conversation). PBS's coverage of that trip noted both sides were explicitly trying to avoid a repeat of the spring's 145%-tariff, rare-earth-control spiral, describing the October truce as still 'fragile.' None of the five fronts closed. The September date itself — 24 September, at the White House — was floated by Trump directly and reported by Transport Topics on 2 September, with reporting elsewhere describing the run-up to it as a 'low-expectations summit' focused on managing the standoff rather than resolving it.

Council on Foreign Relations analysts Rush Doshi, Chris McGuire, Heidi Crebo-Rediker, David Sacks and David Hart argue China enters these meetings with the stronger hand specifically because of what happened in 2025: Xi 'successfully resisted Trump's tariff escalation by threatening to restrict rare earth supplies,' demonstrating that Beijing can compel American concessions without a shot fired. That is a card worth keeping rather than spending — hence a pause, not a permanent stand-down, with a hard expiry date. On the American side, PBS's May 2026 reporting on the Beijing trip describes Trump pressing Xi, without success, to use China's position as Iran's largest oil buyer to help end the US-Israeli war then still running, alongside the standard trade asks — Boeing orders, agricultural purchases — that Washington has sought at every meeting in this sequence. MercoPress's account of the same summit frames the whole five-front agenda as items neither side is ready to resolve outright: Taiwan is a 'core interest' Beijing won't trade away, rare earths are a lever China won't fully surrender, and AI safety talks are, in the reporting's own words, unlikely to produce 'meaningful dialogue.' Likely rather than confirmed: the pattern — truce, pause, summit, no resolution, repeat — is documented across three meetings, but no official on either side has stated the strategy of managed non-resolution as a stated policy; it is this analysis's reading of a repeated pattern, not a claim either government has made about itself.

CFR's analysts are direct about the imbalance: US and allied efforts to build alternative critical-mineral supply chains 'face years, if not decades, to realize,' while China's processing dominance is usable leverage today — an asymmetry that doesn't disappear because a truce paused its use. Xi's stated aims for the September meeting, per the run-up reporting, are to have the summit validate China's standing as a US-level power and to lock in 'tariff predictability' on terms Beijing can live with — both achievable without China giving up anything it would miss. On the US side, the November 2025 fact sheet is a genuine list of deliverables Washington can point to: a specific, falling fentanyl tariff rate, specific tonnages of agricultural purchases, specific commodities exempted from Chinese retaliation. Whether those deliverables represent durable market access or a purchase schedule China can quietly under-deliver on — as it has with past soybean commitments in other administrations — is not something the reporting resolves either way. Uncertain, because 'benefits' here means different things on each side — structural leverage for Beijing, an itemized scoreboard for Washington — and no source compares the two on a common scale.

The structural problem is that every concession in this relationship carries an expiry date rather than a repeal. The reciprocal-tariff pause lapses 10 November 2026; the rare-earth licensing pause lapses 27 November. Businesses on both sides that adjusted contracts, pricing or sourcing around the October 2025 truce are planning around dates that require a fourth renegotiation to survive, not a settled policy. TIME's October 2025 analysis, published before the Busan meeting even concluded, argued that 'sticking to a deal doesn't seem to be Trump's game' — citing 28 separate tariff reversals between his April 2025 'Liberation Day' announcement and mid-July, plus a abruptly terminated Canada trade talk and a snap 10% tariff hike over an advertising dispute — and extending the same skepticism to whatever comes out of a Trump-Xi meeting specifically. Even a nominal ally fared no better: Japan secured an auto-tariff cut from 27.5% to 15% in exchange for a $550bn investment pledge, only for Trump to later reserve the right to reopen it. Likely, not confirmed, because the specific outcome of the 24 September meeting had not happened as this analysis was written — the pattern of reversal is documented across many prior deals, but this one has not yet had the chance to be broken.