目录
TL;DR
1. What Exactly Did WuXi AppTec Win?
2. A Two-Year Timeline: WuXi AppTec Entered the Process Early, but the Full Rationale Emerged Only in Litigation
3. What Is 1260H? A List Becoming the Interface for Multiple Regulatory Regimes
4. Why All 3 of the Pentagon’s Evidentiary Chains Failed
Ground 1: Treating “5.32% of the Fund’s Net Assets” as “5.32% Ownership of the Company”
Ground 2: Recasting a “Third-Party Laboratory” as a Research Partner of Universities
Ground 3: Recasting a Novo Nordisk-Sponsored Study as Collaboration Between WuXi AppTec and a PLA Hospital
5. Why “Heightened Deference on National Security” Did Not Save the Decision
6. Why the Court Considered the Commercial Harm Irreparable
7. Comparing Xiaomi, Luokung, Hesai, and DJI: This Is Not Evidence That “Chinese Companies Generally Win More Easily”
8. What Chinese Companies Can—and Cannot—Replicate
9. What Comes Next: Four Scenarios That Matter
Conclusion: A Victory for Evidentiary Discipline, Not the End of the Policy Cycle
本内容基于公开资料和研报数据整理，不构成任何投资建议，不代表任何个人观点，仅供学习参考，请理性阅读
The August 7 ruling temporarily shields WuXi AppTec from the effects of its 1260H listing, but it changes the evidentiary structure of this case—not the direction of US biotechnology “de-risking” policy.
TL;DR
WuXi AppTec won a preliminary injunction, not a final judgment. The US District Court for the District of Columbia barred the Department of Defense from enforcing this listing and declined to automatically stay the injunction pending appeal. The Pentagon may still appeal or relist the company based on accurate facts and a complete rationale.
The Pentagon lost because all 3 parts of the administrative record mischaracterized the underlying materials. The 5.32% figure represented WuXi AppTec shares as a percentage of the fund’s net assets—not a 5.32% ownership interest in WuXi AppTec. Materials covering 6 clinical studies showed that WuXi AppTec was merely a third-party laboratory, not the applicant, and did not show that it “conducted research jointly” with the relevant universities or PLA hospital.
The court did not undertake substantive review of the national-security determination. It acknowledged the Pentagon’s broad discretion, but held that the Administrative Procedure Act does not allow courts to repair factual errors or supply missing reasoning on an agency’s behalf. The ruling specifically stressed that it had not found WuXi AppTec “incapable of posing a security risk.”
Compared with the Hesai and DJI cases, the dividing line is not whether the company is Chinese, but whether at least 1 independently valid statutory pathway remains. Hesai and DJI both lost under the current 1260H framework because alternative grounds, including designation as a “military-civil fusion contributor,” remained supported by evidence. In WuXi AppTec’s case, all 3 grounds cited by the Pentagon failed.
Investors must track 3 clocks simultaneously: judicial, regulatory, and commercial. The injunction can freeze the government label, but it cannot immediately bring back customer projects that have already moved elsewhere. Outside 1260H, the Office of Management and Budget may still establish a list of “biotechnology companies of concern” under BIOSECURE’s independent criteria.
1. What Exactly Did WuXi AppTec Win?
On August 7, 2026, Chief Judge James E. Boasberg of the US District Court for the District of Columbia issued a preliminary injunction barring the Department of Defense and related defendants from “enforcing, implementing, or otherwise giving effect to WuXi AppTec’s 1260H designation.” The court also denied the Pentagon’s request for a stay pending appeal and required WuXi AppTec to post only a US$1 bond. The immediate effect is to freeze this particular listing while litigation continues—not to invalidate Section 1260H or permanently remove the company from the list (court order, ECF 24).
WuXi AppTec satisfied all 4 requirements for a preliminary injunction: a strong likelihood of success on its Administrative Procedure Act claim; ongoing irreparable commercial and reputational harm from the listing; a balance of equities favoring relief; and a public interest better served by an accurate list compiled in accordance with law. The court did not proceed to adjudicate the complaint’s other claims involving procedural due process, ultra vires action, and statutory vagueness because the Administrative Procedure Act claim alleging “arbitrary and capricious” agency action was sufficient to support relief (court opinion, ECF 25).
The precise interpretation of the victory is therefore this: WuXi AppTec demonstrated that this particular Pentagon listing was likely unlawful and secured temporary relief preventing further harm. It has not yet shown that it could not be listed in the future under different evidence or another statutory pathway.
2. A Two-Year Timeline: WuXi AppTec Entered the Process Early, but the Full Rationale Emerged Only in Litigation
WuXi AppTec’s engagement with the Pentagon did not begin after its listing. On August 26, 2024, the company briefed the Department of Defense on its business, governance, and ownership structure, while denying any relationship with the Chinese military that would warrant designation. It subsequently answered a series of questions in writing. WuXi AppTec was not included on the list published on January 7, 2025.
In late 2024, Congress expanded the 1260H definition by adding “affiliated with” as a qualifying relationship. WuXi AppTec therefore contacted the Pentagon again on August 25, 2025. A meeting initially scheduled for October was postponed until November because of a lapse in appropriations. By October 7, however, Deputy Secretary of Defense Steve Feinberg had already told Congress that WuXi AppTec and other companies should be added to the list. Unaware of that letter, WuXi AppTec again explained its business, ownership, and governance at the November meeting.
On February 13, 2026, WuXi AppTec appeared on a prepublication list accompanied by a single-sentence rationale: “WuXi AppTec is indirectly owned by SASAC and indirectly affiliated with SASTIND and the Chinese People’s Liberation Army.” Approximately 1 hour later, the Federal Register withdrew the document at the Pentagon’s request. WuXi AppTec filed a Freedom of Information Act request for the rationale on May 9 but received no substantive response within the statutory deadline. On May 29, an internal Pentagon report established 3 independent grounds for designation. WuXi AppTec sent a point-by-point rebuttal on June 8; just hours later, the Pentagon nevertheless published the list, which appeared in the Federal Register on June 10 with exactly the same rationale as in February. The company sued on June 11, oral arguments were held on July 22, and the preliminary injunction was granted on August 7 (Federal Register listing notice dated June 10, 2026).
This timeline highlights an underappreciated procedural issue. WuXi AppTec had submitted explanations well in advance, yet the only public rationale available to the company remained a single sentence. The May 29 internal report containing the actual evidence and reasoning was not fully disclosed until litigation began and the administrative record was filed. Administrative-law review ultimately asks whether the decision-making agency considered the company’s explanations, read the evidence accurately, and documented a reviewable causal chain when making its decision.
3. What Is 1260H? A List Becoming the Interface for Multiple Regulatory Regimes
Section 1260H originally required the US Secretary of Defense to identify “Chinese military companies operating directly or indirectly in the United States.” Following legislative amendments, companies can qualify through multiple pathways—for example, if they are directly or indirectly owned or controlled by the Chinese People’s Liberation Army or specified state agencies; have “close formal or informal ties” to those entities; act as their agents; or are identified as military-civil fusion contributors. The law also requires the Pentagon to use the latest information, submit the list to Congress, and publish its reasons for each designation.
The list matters not because it freezes all transactions at once like a traditional sanctions list, but because it has become a “base data layer” referenced by other rules. Defense procurement restrictions, certain lobbying restrictions, and some Department of Homeland Security and Department of Energy rules incorporate 1260H status. More importantly, Section 851 of the US National Defense Authorization Act for Fiscal Year 2026, effective December 18, 2025, established the BIOSECURE framework. A company involved in biotechnology equipment or services that is also listed under 1260H may enter the Category A pathway for designation as a “biotechnology company of concern.” The Office of Management and Budget may also use the independent Category B pathway based on foreign-adversary governance, control or agency relationships, biotechnology activities, and national-security risks (Section 851 of Public Law 119-60).
The two concepts must not be conflated. Inclusion on the 1260H list does not mean that BIOSECURE procurement restrictions automatically take full effect that same day. The Office of Management and Budget must publish its list of companies of concern within 1 year of enactment, then issue implementation guidance, after which the Federal Acquisition Regulatory Council must amend the Federal Acquisition Regulation. Restrictions for different categories take effect 60 days or 90 days after those amendments, while existing contracts remain subject to statutory transition arrangements. The 1260H listing is an important prerequisite and risk multiplier, but it is not the entire process.
Precisely because the list is becoming an interface across multiple regulatory regimes, the market will act before the legal effective date. Corporate compliance teams are primarily concerned that a multiyear R&D; project might lose eligibility for government contracts, grants, or regulatory filings midway through its life. The commercial damage from a 1260H designation often occurs before the formal restrictions take effect.
4. Why All 3 of the Pentagon’s Evidentiary Chains Failed
The Pentagon’s internal report stated that its 3 grounds were independent and that any 1 would be sufficient to support the listing. The court reviewed each ground under that structure. The problem was not merely that the evidence carried insufficient weight; all 3 grounds described the underlying sources as showing something they did not.
Ground 1: Treating “5.32% of the Fund’s Net Assets” as “5.32% Ownership of the Company”
The Pentagon asserted that a military-civil fusion investment fund controlled by Aviation Industry Corporation of China held a 5.32% interest in WuXi AppTec. Because Aviation Industry Corporation of China is supervised by the State-owned Assets Supervision and Administration Commission of the State Council, the Pentagon concluded that WuXi AppTec was indirectly owned by SASAC. The problem was that the 2 cited articles said WuXi AppTec shares accounted for 5.32% of the fund’s net assets—the portfolio weight—not that the fund owned 5.32% of WuXi AppTec’s total equity.
This was not a disagreement over valuation methodology; it was the wrong denominator. A household investing 5% of its savings in a blue-chip stock does not thereby own 5% of the listed company. WuXi AppTec estimated during litigation that a genuine 5.32% equity stake would have been worth approximately US$850 million in 2019, more than 40 times the fund’s total year-end assets. The position described in the source materials was worth approximately US$200,000 and represented approximately 0.001% of WuXi AppTec’s equity. The court specifically noted that the latter 2 figures came from outside the administrative record and therefore did not need to be accepted. The Pentagon’s conclusion already failed on the basic point that a stock’s share of fund assets is not the fund’s share of a company’s equity.
More seriously, the Pentagon interpreted 2 articles from 2019 and 2024 as describing separate investments, when the later article was actually discussing the same 2019 position. The court therefore found that the agency was not choosing among competing reasonable interpretations but had “mischaracterized the evidence.”
Ground 2: Recasting a “Third-Party Laboratory” as a Research Partner of Universities
The Pentagon asserted that WuXi AppTec conducted research jointly with 5 universities supervised by the State Administration of Science, Technology and Industry for National Defense and had received approval from the Ministry of Science and Technology. Its sole source was a 2022 compilation of approvals for clinical trials involving human genetic resources. The table divided participants into 4 categories: applicants, medical institutions, contract research organizations, and third-party laboratories. The applicants were primarily European and US multinational pharmaceutical companies, while the medical institutions were hospitals affiliated with the relevant universities. WuXi AppTec or its subsidiaries appeared only in the rightmost column labeled “third-party laboratory.”
The table did not identify WuXi AppTec as an applicant or describe any agreement, joint project, or collaborative relationship between WuXi AppTec and the universities. Relying on the same table, the Pentagon recast “providing testing services” as “obtaining approval and conducting research jointly with universities,” without explaining why this more remote role constituted the “close formal or informal ties” required by law. The court also observed that the University of Michigan appeared as an applicant in the same compilation, with a Peking University-affiliated hospital serving as the research institution. If a third-party laboratory’s remote involvement were sufficient to establish a close affiliation, the more direct applicant relationship would push the Pentagon’s interpretation to an absurd boundary.
Ground 3: Recasting a Novo Nordisk-Sponsored Study as Collaboration Between WuXi AppTec and a PLA Hospital
The alleged PLA affiliation arose from a study in the same approval compilation. The table showed Novo Nordisk as the applicant, the PLA General Hospital as the research institution, and WuXi AppTec’s Shanghai subsidiary as merely 1 of 3 third-party laboratories. The Pentagon nevertheless asserted that WuXi AppTec had obtained approval and conducted drug research with the PLA hospital. Because this was the same role-identification error found in the preceding 5 studies, the third ground failed as well.
The court also questioned whether “directly or indirectly” modifies “affiliated with” in the current statutory text. Grammatically, the phrase more naturally modifies only ownership, control, and beneficial ownership. “Affiliated with” is already defined as having close formal or informal ties, so adding “indirect affiliation” may be redundant. The court did not resolve this interpretive dispute because all 3 factual chains failed even under the Pentagon’s broader reading.
The case therefore sends an important legal signal: the court did not require the Pentagon to satisfy a criminal-conviction standard of proof. It merely required the agency to state accurately what the materials showed, what role the company played, and why that role fell within a statutory relationship. The Pentagon retains national-security discretion at all 3 steps, but it cannot skip any of them.
5. Why “Heightened Deference on National Security” Did Not Save the Decision
U.S. courts generally defer heavily to executive judgments on foreign affairs and national security, and 1260H cases are no exception. In opposing the motion, the Department of Defense argued that judicial review need only confirm the existence of “substantial evidence,” allowing an agency to draw reasonable inferences from uncertain materials. Judge Boasberg accepted that low threshold but stressed that it was not no threshold at all: the 5.32% denominator error was not a reasonable inference, nor was recasting a third-party laboratory as an applicant and partner a choice between two permissible interpretations.
The governing administrative-law doctrine is the Chenery principle: a court may review only the reasons the agency gave when making its decision. Government lawyers cannot supply a new rationale during litigation, and the court cannot reconstruct the relationships for the agency and conclude that they were “perhaps close enough.” If a small open-market holding might still constitute indirect ownership, or third-party testing services might still establish a close affiliation, the Department of Defense must first identify the facts accurately, define the applicable threshold, and explain the causal chain.
The court was also willing to intervene because the administrative record contained no classified intelligence cited by the Department of Defense and no ongoing administrative reconsideration that might supplement the record. The court faced the Department’s own 3 stated rationales and the public materials it cited, with direct conflicts on every point. The injunction was also strictly limited to WuXi AppTec’s specific designation: it did not affect other listed companies, prevent the Department from protecting the supply chain, or bar it from redesignating WuXi AppTec based on new evidence. The court therefore framed the public interest in two parts: the United States plainly has an interest in identifying military-linked companies, but it also has an interest in ensuring that a list carrying severe consequences is accurate and lawful.
6. Why the Court Considered the Commercial Harm Irreparable
WuXi AppTec had to show not only that the Department of Defense was likely acting unlawfully, but also that waiting for a final judgment would cause irreparable harm. The court credited concrete evidence—not abstract valuation losses—of customers pausing work, canceling projects, shifting business, and suppliers no longer making deliveries within weeks of the designation. According to the court record, WuXi AppTec has 6 facilities, approximately 450 employees, and more than 1,000 customers in the United States; approximately 70% of its 2025 revenue came from U.S. customers. Some customers hold federal contracts or depend on federal funding, giving them a rational incentive to change suppliers before the BIOSECURE Act formally takes effect.
The stickiness of the CRDMO business makes this harm difficult to reverse. A drug program can take years to move from R&D; through process validation and analytical-method transfer into commercial production. Changing service providers often requires revalidating manufacturing processes, transferring testing methods, and submitting supplemental filings to the U.S. Food and Drug Administration. Once customers incur those switching costs, they are unlikely to return even if WuXi AppTec ultimately prevails and is removed from the list. The loss extends beyond current contracts to follow-on orders as projects advance and collaborations on next-generation products.
The government argued that the initially affected projects remained small relative to WuXi AppTec’s approximately US$6.7 billion in annual global revenue and that many customers were merely “waiting and watching.” The court declined to treat losses from the first few weeks as a ceiling: the harm was expanding, and U.S. customers sit at the core of the company’s business. Sovereign immunity also means that even if WuXi AppTec ultimately wins, it generally cannot recover its losses from the U.S. government. The combination of unrecoverable financial losses, relationships that are difficult to rebuild, and continuing project migration gave the preliminary injunction its urgency.
This is also central to understanding the market reaction. The judicial clock advances through hearings and rulings; the regulatory clock through lists, guidance, and procurement rules; and the commercial clock through customer project cycles and risk committees. The 3 clocks do not move in sync. The injunction can require the government to no longer give legal effect to this designation, but it cannot order private customers to resume business.
7. Comparing Xiaomi, Luokung, Hesai, and DJI: This Is Not Evidence That “Chinese Companies Generally Win More Easily”
It is tempting to describe the WuXi AppTec case as a new template for Chinese companies challenging U.S. government lists, but adjacent precedents support a more measured conclusion.
The Xiaomi and Luokung cases are instructive on reputational harm and the limits of affiliation, but they applied the former Section 1237, which has since been repealed, and therefore do not directly determine outcomes under the current 1260H framework (Xiaomi and Luokung Technology rulings).
Hesai and DJI are more comparable. In Hesai, the court held that the company’s location in or affiliation with a military-civil fusion enterprise zone, together with LiDAR’s substantial military applications, was sufficient to support its classification as a “military-civil fusion contributor” (Hesai ruling). DJI is particularly instructive: the court rejected the Department of Defense’s view that merely holding shares amounted to ownership and also rejected several other rationales. It nevertheless found one valid pathway based on support from a national enterprise technology center and the significant military uses of drone technology, allowing the full designation to stand (DJI ruling).
These cases establish the real logic of victory and defeat. A challenger need not prove that every word from the Department of Defense is wrong, while the Department need not prevail on every rationale. A designation may survive as long as one independent pathway is supported by the record, rests on complete reasoning, and falls within the statutory definition. WuXi AppTec is unusual because all 3 pathways rested on misread roles or figures, and the Department did not separately assert a “military-civil fusion contributor” pathway as a fallback.
8. What Chinese Companies Can—and Cannot—Replicate
The first replicable step is to build a litigation-ready factual foundation before designation. Ownership evidence should go beyond a shareholder list and distinguish among portfolio weight in a fund, percentage ownership of the issuer, beneficial ownership, voting rights, and actual control. Evidence of collaboration should do more than list projects in which parties appear together; it should distinguish among applicants, research institutions, contract research organizations, third-party laboratories, payers, and contractual counterparties. WuXi AppTec’s decisive tool was not a sweeping policy argument, but restoring the correct denominators and roles from the original tables.
Second, companies should submit evidence while the administrative record is still being formed. Courts generally review only the record available to the decision-maker at the time. Declarations submitted after litigation begins can establish harm but may not be sufficient to overturn the substantive determination. WuXi AppTec had engaged with the government since 2024, explained its governance and ownership twice, and sent a letter rebutting each allegation before the formal designation. This allowed the court to distinguish between “the company did not respond” and “the agency did not address the response.”
Third, companies must challenge every independent rationale. If the Department of Defense identifies 4 pathways—ownership, agency, affiliation, and military-civil fusion—a challenger may still lose after defeating 3. Litigation strategy should begin with a matrix matching each statutory element to each original item of evidence and each inferential leap, then identify which rationale could survive independently. DJI is the clearest counterexample: the failure of some rationales did not change the ultimate outcome.
Fourth, evidence of irreparable harm should be contemporaneous. Customer suspension emails, project-cancellation records, supplier notices, exposure to federal funding, and the validation and filing steps required to replace a supplier are more likely to support an injunction than a falling share price. For software, hardware, or manufacturing companies, analogous evidence may include repeated certification, restarted design-in processes, inventory transitions, removal from cloud platforms, or the loss of long-term framework agreements.
The limits of replication are equally important. Courts are likely to continue deferring heavily to agencies if the administrative record contains undisclosed classified materials, if there is an articulable and substantive connection between the company’s technology and military use, if a military-civil fusion contributor pathway has independent evidentiary support, or if Congress directly expands the definition in new legislation. The WuXi AppTec case does not create a general safe harbor merely because an open-market shareholding is small or because a project serves commercial customers. It establishes only that an agency must describe the actual relationships accurately.
9. What Comes Next: Four Scenarios That Matter
The first scenario is that the current case proceeds to a decision on the merits. A “substantial likelihood of success” at the preliminary-injunction stage does not guarantee an ultimate victory, but the ruling already delivered a sharply negative assessment of three evidentiary chains. Unless the administrative record changes, the Department of Defense will face an uphill battle at the merits stage.
The second scenario is that the Department of Defense appeals and seeks a stay from the federal appellate court. The district court denied an automatic stay, but the appellate court can still make an independent determination. As long as the injunction remains in effect, the legal effect of the current Section 1260H designation is frozen; if a stay is granted, commercial pressure may intensify again.
The third scenario is that the Department of Defense remands the matter for reconsideration or redesignates the company. The ruling expressly preserves this route. The Department could verify the actual shareholding structure, supplement the record on contractual and organizational relationships, introduce new public or classified evidence, and offer different reasoning. Any new decision would create a new administrative record and a new target for litigation. The rejection of the current rationale therefore does not create a permanent safe harbor.
The fourth scenario involves BIOSECURE. Even if the Category A route through Section 1260H is temporarily unavailable, the Office of Management and Budget may still seek to designate a company under the independent Category B criteria: governance, control, or agency relationships involving a foreign adversary; participation in biotechnology equipment or services; and national-security risks arising from factors such as joint research, multi-omics data, or data obtained without consent. This route has its own notice, 90-day response, and review procedures. Whether WuXi AppTec appears on a future list will depend on OMB’s specific reasoning and cannot be determined in advance by the injunction in this case.
For investors, the most useful indicators are not a snapshot showing whether WuXi AppTec is no longer on the list, but four sets of developments: whether the Department of Defense advances a new, independent statutory basis; whether the appellate court changes the status of the injunction; how far the OMB list and Federal Acquisition Regulation process have progressed; and whether customer pauses, new project awards, and previously transferred projects show verifiable signs of recovery. The first three determine institutional risk; the last determines when revenue risk truly ends.
Conclusion: A Victory for Evidentiary Discipline, Not the End of the Policy Cycle
The WuXi AppTec case shifts the Section 1260H dispute away from an abstract US-China policy conflict and back to the most basic requirements of administrative law: the denominator behind a figure must be correct, roles in a table cannot be interchanged, and the agency must articulate its own reasoning connecting the evidence to its conclusion. National-security concerns may make judicial review more deferential, but they cannot transform factual errors into substantial evidence.
The victory is therefore both important and limited. It is important because it shows that companies are not entirely without recourse against designation, and that an accurate administrative record remains a constraint the US government must observe. It is limited because the court did not invalidate the Section 1260H framework, reject US security concerns over biotechnology supply chains, or prevent the government from acting again on a legally sufficient record. For WuXi AppTec, August 7 is better understood as an opportunity to reset the judicial clock. The ultimate outcome will depend on whether regulators rebuild the case, whether customers return, and where the slower but broader BIOSECURE clock ultimately leads.