When Donald Trump sat down with Xi Jinping in Beijing on May 15, neither government's official readout mentioned Taiwan's arms purchases. Trump himself, however, was considerably less discreet.
In a Fox News interview conducted during the trip — and again in remarks to reporters aboard Air Force One on the return flight — Trump confirmed that he and Xi had discussed U.S. arms sales to Taiwan "in great detail." He then volunteered that he was holding the matter "in abeyance," that the decision "depends on China," and, in a phrase that should alarm anyone tracking cross-strait security, that the sales represent "a very good negotiating chip for us."
The remarks triggered immediate concern that Washington might freeze a pending $14 billion arms package — the largest single sale ever proposed to Taiwan. Several foreign outlets have since reported it could be delayed "indefinitely."
That would be consequential in ways that stretch well beyond the dollar figure.
What Is Actually in the Package
The precise contents of the $14 billion package remain partly unconfirmed, but credible accounts describe four major weapons systems, three of which are publicly known.
The first is one additional PAC-3 MSE (Patriot Advanced Capability-3 Missile Segment Enhancement) battery, paired with new LTAMDS radar systems. This would bring Taiwan's total Patriot battalions to four and substantially expand its missile stockpiles.
The second is nine NASAMS (National Advanced Surface-to-Air Missile System) batteries; combined with three units already in procurement, Taiwan's eventual total would reach twelve.
The third is the IBCS — the U.S. military's Integrated Air and Missile Defense Battle Command System, currently in active service with American forces.
These three systems form the essential core of what Taiwan calls "T-Dome" — its active island-wide air defense initiative. The upgraded Patriot battalion would serve as the network node through which IBCS integrates every existing and incoming air defense asset: the current Patriot units, the twelve future NASAMS batteries, American-made counter-drone systems, and U.S.-supplied air defense radars.
Once that foundation is in place, the architecture extends further to Taiwan's domestically produced Sky Bow, Brave Wind, and land-based Sword II missiles, along with indigenous radar platforms. The goal is to connect every air defense missile system and C5ISR (command, control, communications, computers, cyber, intelligence, surveillance, and reconnaissance) node across the island into a single, fully integrated operational network — shifting from a collection of independent batteries into a unified, island-wide defensive architecture.
The military logic is clear: if one battery's radar or Engagement Control Station is destroyed by a PLA missile or drone strike, that battery can still receive targeting data and firing commands from other nodes on the network and remain operational.
The fourth item in the package has not been officially confirmed. However, a publicly released portion of a classified briefing document published by Taiwan's Ministry of National Defense on April 20 — covering planned expenditure under the special arms procurement budget — references a "soft-kill and hard-kill hybrid counter-drone system" with all substantive details redacted, suggesting it is likely part of this sale as well.
How Close the Sale Came to Congress
Understanding what has been frozen requires understanding how far along the process was.
Under standard U.S. arms transfer procedures, the State Department must provide informal advance notice to relevant congressional committees twenty to forty days before a formal Letter of Offer and Acceptance is submitted. Those informal briefings are typically the source from which journalists first learn the details of a pending sale.
Foreign media first reported the existence of this $14 billion package on February 6 — just two days after Xi Jinping, during a February 4 phone call with Trump, explicitly urged the United States to "handle arms sales to Taiwan with great caution." The fact that reporters had already been briefed in enough detail to publish specifics on February 6 strongly suggests that senior congressional members had received informal notification in mid-January.
That, in turn, means the package had already cleared the interagency review process, and the administration had been on track to submit the formal draft offer to Congress before the end of February — well ahead of Trump's then-planned state visit to Beijing in late March or early April, which would have made the sale a done deal before any summit.
In other words, the administration almost certainly halted the sale at the final stage — one step short of congressional submission — following Trump's February 4 call with Xi, on what was effectively a direct instruction from Trump and his inner circle.
Trump may yet approve the sale in the near term. The more likely scenarios, however, are that Washington delays the entire package until after the U.S. midterm elections on November 3 — or even into early next year — or adopts a compromise approach: approving a small portion now and holding the rest until after the midterms or beyond.
Outright cancellation remains unlikely. Because informal congressional briefings have already taken place, abandoning the sale entirely would provoke significant backlash on Capitol Hill. It would also leave governments across the Indo-Pacific with the clear impression that Trump had yielded to Xi — a reputational cost the White House would struggle to absorb.
Why Trump Is Hesitating
Several forces are pushing Trump toward delay. The failure of his reciprocal tariff strategy and the ongoing Iran conflict have weakened Washington's negotiating leverage.
More critically, Beijing has read the political situation with precision. Trump's domestic approval ratings have fallen partly because of the Iran war, and the erosion has reached into the Republican base. Under these conditions, whether Trump can return from Beijing with a substantial economic package has direct implications for Republican prospects in the midterms.
Xi has reportedly agreed to visit Washington on September 24, with three additional Trump-Xi meetings expected before the end of the year. This calendar gives Beijing a powerful structural tool: it can release its economic commitments in stages — making promises incrementally and delivering them selectively — to sustain influence over U.S.-Taiwan political and military relations, including arms sales, for the remainder of the year.
The Chinese purchasing list that has emerged from the Beijing summit includes 200 Boeing commercial aircraft (with up to 750 if initial orders prove satisfactory), as many as 450 aircraft engines, oil and liquefied natural gas in unspecified quantities, at least $17 billion in U.S. agricultural products annually through 2028, and the reopening of American beef and poultry imports.
Of these commitments, only the Boeing order and the beef reopening are sufficiently concrete. The remainder — energy, LNG, and agricultural products — are expressions of intent that require further rounds of negotiation before they become actual contracts.
Trump's push for Xi to visit Washington on September 24 reflects the calculation that Xi, if he agrees to come, might use the occasion to formalize and expand those commitments as a diplomatic gesture — and that locking in that timing gives the economic effects enough runway to influence voter sentiment before November 3.
The geographic overlap between these purchases and Republican electoral territory is not incidental. Aircraft, engines, agricultural goods, and energy all map onto states and districts that form the Republican base — constituencies already showing signs of strain over the Iran war.
Republicans currently hold 220 of 435 seats in the House of Representatives against the Democrats' 215, a margin of just five seats above the 218 needed for a majority. If Republicans lose three seats in districts they currently consider safe, they lose control of the chamber — and with it, Trump's ability to govern with a fully unified government.
The manufacturers involved — Boeing, aircraft engine producers, and energy companies — also wield considerable influence over campaign fundraising for both parties. Three years of agricultural purchases alone would total $51 billion, roughly 3.6 times the value of the $14 billion arms package. To protect the pathway from Chinese "intent" to "signed contract," these industries may have little appetite for Trump approving the Taiwan sale before year's end.
Most significantly, Xi's February 4 phone call with Trump and his May 14 discussion of the arms sale "in great detail" have effectively tied the question of whether the sale proceeds to Xi's personal prestige. If Trump approves this record-breaking package shortly after returning to Washington — or at any point during the remaining Trump-Xi meetings this year — Beijing, to protect Xi's standing, may cancel all subsequent summits and rescind its purchasing commitments. The electoral damage that would inflict on the Republican Party is not a scenario Trump or his senior advisers can afford to dismiss.
What Delay Actually Means for Taiwan's Defense
If Trump defers the sale until November or early next year, the paper delay is five to seven months. The delay Taiwan's military would actually experience in receiving the equipment is likely to be far longer.
U.S. defense industrial capacity was already severely strained before the Iran conflict. It has since deteriorated further. According to reports from late April, Washington has already warned European allies — including the United Kingdom, Poland, Lithuania, and Estonia — to prepare for longer delays in American weapons deliveries. Among the systems specifically flagged were NASAMS, which features in the $14 billion Taiwan package, and HIMARS (High Mobility Artillery Rocket Systems), which the U.S. has already separately approved for sale to Taiwan.
The PAC-3 MSE situation is particularly acute. Before the Iran war, the U.S. Army had already announced in September 2025 that it would procure nearly 2,000 PAC-3 MSE missiles and activate four additional Patriot battalions of its own. In January 2026, Washington announced the sale of 730 PAC-3 MSE missiles to Saudi Arabia.
Since the outbreak of the Iran conflict, U.S. forces have expended approximately 1,430 Patriot missiles — roughly 60 percent of pre-war stockpiles. Even assuming no upward revision to combat reserve requirements, American forces alone need close to 3,500 missiles to restore inventory, nearly six times the annual production rate of 600 missiles as of 2025. American forces hold first priority in production allocation.
On top of that, Israel, Saudi Arabia, Qatar, Bahrain, and the United Arab Emirates are all likely to submit new PAC-3 MSE procurement requests to replenish wartime consumption.
Every additional month that passes before the Taiwan sale is formally approved allows other buyers — Middle Eastern states, and NATO members seeking enhanced ballistic missile defense — to lock in earlier production slots. A five-to-seven-month approval delay does not produce a five-to-seven-month delivery delay. It produces a substantially longer wait, as Taiwan's order is pushed further back in a production queue that is already severely congested.
Because the PAC-3 MSE system plays the pivotal role in T-Dome — serving as the integration node through which IBCS connects all other components — a significant slippage in its delivery timeline threatens to push back the date by which Taiwan can achieve full, island-wide integration of all air defense and C5ISR systems well beyond 2033, and potentially beyond 2035.
That last date matters. The year 2035 is the horizon by which the PLA's "Joint Fires Strike" operational capability is expected to reach a qualitatively higher level of threat. A Taiwan that has not completed its integrated air defense network by that date will face that threat at a severe structural disadvantage — a disadvantage that a delayed arms approval, compounding through an already congested production queue, may make very difficult to overcome.
*The author is a Research Fellow at the Center for Integrated Strategy and Technology Studies, Tamkang University.
