Trump Invokes WWII Alliance as He Brings Xi Back to the Table
Trump and Xi meet at the White House for the first time in four months, with AI cooperation and midterm-era concessions on the agenda. The wartime alliance framing signals a managed rivalry — but the real story is how technology governance is becoming the new battleground.
A Summit Disguised as a Homecoming
Donald Trump greeted Xi Jinping at the White House on September 24 with language that would have sounded bizarre a decade ago: he called China an ally from the Second World War, a partner whose shared interests could bridge differences. The remark was not historical scholarship. It was political theater — and the kind that reveals exactly where Washington and Beijing both want the relationship to go.
The meeting marks the first face-to-face encounter between the two leaders since May, when they held two days of talks in Beijing that ended with Xi walking Trump through the Temple of Heaven. Four months of silence since then has been filled with tariffs, export controls, and the usual friction over Taiwan and Iran. The fact that they are sitting down again says less about resolution than about management.
What made the September 24 greeting notable was not just the content but the staging. Trump stood at the top of the South Portico steps, flanked by his Cabinet and military advisors, while Xi arrived with a delegation that included Liu He — a reminder that China still views trade and economic coordination as the backbone of its diplomacy. The photograph of the two leaders shaking hands, with the Capitol dome rising behind them, will be the one circulated in capitals around the world. But the substance of what followed revealed a narrower, more pragmatic agenda than the rhetoric suggested.
The AI Quiet Question
What English-language coverage has largely underreported is how central artificial intelligence has become to this round of talks. According to analysis by Tokyo University professor Makoto Kawaishima, the two sides are structuring a new domain of dialogue around AI — a sector where the competition is fiercest but also where the costs of pure confrontation are highest. AI safety, model governance, and export controls on advanced chips sit at the intersection of both economies.
The framing matters because it signals something about how Trump wants to sell this to his base. Pairing AI cooperation with the wartime alliance rhetoric creates a narrative arc: the two countries that defeated fascism together can now steer the most powerful technology humanity has ever built, rather than letting it become a weapon of mutual destruction. Whether that narrative holds up against the reality of competing chip bans, Chinese restrictions on critical minerals, and US efforts to tighten alliance coordination around semiconductor supply chains is another question entirely.
Kawaishima points out that the working-level discussions have already produced a draft framework for bilateral AI oversight consultation — not an agreement, but a structure that could survive political shifts. This is the institutional scaffolding both sides have been building quietly for months. The public optics are about Trump and Xi at a formal dinner; the real outcome will be measured in whether ministerial working groups on model transparency and chip-export compliance can produce measurable deliverables by year’s end.
The stakes extend beyond bilateral relations. A US-China framework on AI governance would set de facto global norms, since both countries together account for the majority of the world’s advanced compute capacity and large-language-model development. Whatever rules emerge from Washington and Beijing will become the template that the EU, India, and others either adopt or contend with.
The Midterm Calculus
Kawaishima’s analysis points to a practical dimension that often gets lost in summit theatrics: China is likely to offer economic concessions ahead of the US midterm elections. Purchases of American agricultural and industrial goods are the kind of transactional gesture that can be spun domestically as diplomatic leverage. It is also the kind of gesture that gives Trump a tangible to hold up in swing-state corridors.
This is not new playbook. It is the same dance that has played out in earlier cycles of US-China negotiation — China buys, the US pretends the deal was earned rather than structured. What is different this time is the AI overlay. Concessions on tech access and research collaboration carry far higher strategic stakes than soybeans, and both sides know it.
The midterms add urgency. Congressional Democrats are already preparing talking points around US-China tech dependence, and any perception that Trump is conceding ground on semiconductors or AI will be seized upon. At the same time, Republican lawmakers in energy and agriculture states are incentivized to welcome the economic deals. The summit’s timing — three months before Election Day — is no accident. It places the leaders in a moment where both can claim wins without yet facing the full scrutiny of a post-summit policy review.
Who Wins, Who Loses
The immediate winner is the institutional architecture around US-China dialogue itself. As Kawaishima notes, ministerial-level working groups on trade and economic issues have continued to function even while headlines focused on confrontation. The September 24 meeting is partly an exhibition of those quiet channels — proof that the relationship has plumbing, not just pressure valves.
But the longer-term stakes run deeper. Japan and South Korea are watching closely. If Washington and Beijing establish a bilateral AI governance framework, it could reshape the rules that Tokyo and Seoul have to follow within their own technology sectors. Both nations already face a difficult balancing act: they host critical semiconductor manufacturing capacity and depend on US security guarantees, but they also carry deep trade ties with China. A US-China tech framework that excludes them from meaningful consultation would force an uncomfortable choice. One attended as partners might let them claim a seat at the table.
China’s positioning on Iran also carries implications: Beijing is expected to take a pro-Tehran stance at the talks, but not at the cost of rupturing the broader relationship. That calculus matters for allies who have their own calculations about Iranian influence in the region. Gulf states, Israel, and European capitals all read Beijing’s moves as indicators of whether China is willing to bear costs for its partnerships — and this summit will provide new data points.
Taiwan remains the unspoken third rail. No summit of this magnitude can avoid it, yet neither side has an incentive to force a confrontation over it right now. The managing-rivalry posture means Taiwan becomes a background risk rather than a foreground crisis — temporarily. That is not stability. It is postponement.
What Comes Next
The evening banquet, attended by Elon Musk and other technology figures, underscores how personal the tech angle has become. This is not just about state-to-state policy. It is about which ecosystem — American or Chinese — will set the norms for the next generation of AI infrastructure.
If the summit produces a joint statement on AI governance or a restart of semiconductor dialogues, it will be framed as a breakthrough. If it produces nothing concrete, it will be read as another cycle of performative engagement. The difference between those outcomes will depend on whether the working-level conversations Kawaishima describes actually produced something worth displaying — or whether the speeches were the product all along.
For allies across Asia, the message is clear: Washington is keeping its direct line to Beijing open, and it is doing so on terms that center technology as much as trade. The question is whether that openness comes with guardrails for partners who find themselves caught between the two powers.
The alliance rhetoric is nostalgic. The AI agenda is not. Both are part of the same strategy — and neither tells the whole story. What this summit ultimately reveals is that the US-China relationship has settled into a new rhythm: confrontation without rupture, cooperation without trust, and a technology race where the finish line keeps moving. The question for the coming years will not be whether these two powers can get along. It will be who writes the rules of the systems they both depend on.