The debate behind The Pentagon’s Consumer-Tech Blacklist Has a Point, and a Problem

Washington is justified in treating BYD, Alibaba and Baidu as national-security risks, not because each is proven to be a weapons producer, but because Section 1260H is explicitly designed for the harder problem: civilian-looking firms that can support China’s defense industrial base through military-civil fusion. A Pentagon 1260H list designation means the Department of Defense has identified a Chinese company operating directly or indirectly in the United States as either tied to specified Chinese military/security/state-industrial organs or as a military-civil fusion contributor while providing commercial services, manufacturing, producing or exporting. The statutory definition includes entities affiliated with MIIT, SASTIND, SASAC, MSS or the PLA, entities receiving Chinese state assistance through defense-industrial planning, entities in MCF zones, entities with military production licenses, and entities advertising on Chinese military procurement platforms. That text defeats the narrow rebuttal that “consumer companies are not defense companies.” Congress deliberately made dual-use contribution, not arms sales, the threshold.

The strongest evidence is sectoral and company-specific. First, China’s MCF system is not hypothetical. CSET’s 2025 study of 2,857 PLA AI-related contract notices from 2023-2024 found that nontraditional vendors with no self-reported state ownership won 764 contracts, and that civilian AI vendors and research institutions are active participants in PLA procurement. That supports the policy logic for screening AI, cloud, mapping, battery and vehicle platforms before they become embedded in U.S. defense supply chains. Second, Alibaba is not merely an e-commerce firm for this purpose. Public U.S. congressional material alleges that Alibaba Cloud signed a 2017 military-civilian fusion cloud cooperation agreement with Norinco, a major state defense company, and a cooperative agreement with the National University of Defense Technology; it also describes Alibaba Cloud participation in a PLA Academy of Military Sciences Military Big Data Forum. AP reports that, in the 2026 designation, the Pentagon cited Alibaba’s affiliation with MIIT as relevant. Alibaba denies being a military company, but denial does not erase the policy relevance of cloud infrastructure, data access and AI services to defense modernization.

Third, Baidu shows the dual-use problem clearly. AP reported that a PLA Information Engineering University paper tested Baidu’s Ernie model to generate simulated military response plans; Baidu denied any partnership and said Ernie was publicly available. That is a fair caveat, but it still illustrates why frontier AI platforms are strategically relevant even absent a bespoke weapons contract. BYD’s case is similar in kind: EVs, batteries, power electronics, fleet data and autonomous-vehicle systems are militarily relevant industrial inputs, and public Senate material alleges BYD’s cooperation with PLA-linked weapons-base and MCF activity. The practical bite is real: effective June 30, 2026, DoD is barred from new, renewed or extended contracts for goods, services or technology with 1260H entities or affiliates, and in 2027 the bar extends to procurement including goods or services produced or developed by listed entities. That matters to investors and suppliers because 1260H is an early-warning signal for procurement exclusion, export-control scrutiny, delisting pressure and future sanctions risk.

Advocate B’s strongest point is that the public record is thinner than the policy label sounds: AP reports that the Pentagon’s stated basis for Alibaba, BYD and Baidu prominently includes affiliation with MIIT, and the government has not publicly released a detailed evidentiary dossier showing each firm’s direct PLA contracts. That point has genuine merit because a national-security list should not become an opaque shortcut for branding any large Chinese industrial firm as military-linked merely because it exists under Chinese industrial policy. It also matters that 1260H is not, by itself, a full blocking-sanctions regime.

But that caveat does not defeat the justification for the designation. The relevant standard is not criminal conviction or proof of direct weapons sales; it is risk-based identification under a statute Congress deliberately wrote around military-civil fusion. Section 1260H covers commercial-service, manufacturing, producing and exporting entities that are affiliated with MIIT, SASAC, SASTIND, MSS or the PLA, and separately covers military-civil fusion contributors receiving state assistance through defense-industrial planning, participating in research partnerships, residing in MCF zones, holding military-production credentials, or appearing on military procurement platforms. In other words, B is right that the aperture is broad, but wrong to treat breadth as a defect when the adversary strategy being addressed is precisely the conversion of civilian industrial capacity into military capability.

The company-specific evidence also goes beyond a generic “China ecosystem” inference, especially for Alibaba. A 2023 Senate letter alleged that Alibaba Cloud signed a 2017 MCF cloud cooperation agreement with Norinco and a cooperative agreement with the National University of Defense Technology, and that an Alibaba Cloud executive participated in a PLA Academy of Military Sciences Military Big Data Forum about transferring e-commerce technologies to national defense. Those allegations are not final adjudications, but they are exactly the kind of cloud, data and defense-industrial nexus that 1260H is meant to flag.

For Baidu, AP’s report that it is an AI and self-driving company affiliated with MIIT should be evaluated against broader empirical evidence that the PLA is buying AI from civilian and nontraditional suppliers. CSET’s review of 2,857 PLA AI-related award notices from 2023-2024 found a consequential role for nontraditional vendors and research institutions. That does not prove Baidu’s specific culpability, but it makes Washington’s caution toward major Chinese AI models, mapping and autonomy platforms reasonable rather than speculative. BYD is the least publicly detailed case, but EV batteries, power electronics, autonomous systems and fleet data are dual-use industrial inputs; when coupled with MIIT affiliation under a statute aimed at defense-industrial contribution, early-warning treatment is defensible.

Finally, B understates the practical bite. Crowell’s analysis explains that direct and indirect DoD procurement restrictions tied to 1260H take effect on June 30, 2026 and June 30, 2027, and that screening providers already incorporate the list, causing transaction delays and compliance reviews. Those consequences are not accidental overreach; they are the point of a prudent supply-chain warning system.

B’s strongest argument is that Washington has not publicly shown enough firm-specific evidence to justify attaching a military-linked label to three globally traded consumer-technology firms. Section 1260H is broad: it reaches informal affiliation with MIIT, military-civil fusion zones, industrial-policy assistance, military procurement advertising, and a catch-all Defense Secretary determination, so a designation can look like a policy choice to accelerate decoupling rather than a demonstrated security finding. B is also right that 1260H is not a general sanctions regime by itself, yet its reputational and compliance consequences can become much broader than the formal defense-contract ban.

I still think A’s position is stronger because B implicitly applies too high a burden of proof for a supply-chain warning list. The legal and policy question is not whether BYD, Alibaba or Baidu have been publicly proven to sell weapons to the PLA; it is whether their technologies and institutional links create a material risk of supporting China’s defense industrial base. Congress wrote Section 1260H to cover commercial-service, manufacturing, producing and exporting firms that are affiliated with specified Chinese state-industrial or security organs, including MIIT, or that qualify as military-civil fusion contributors. That statutory structure reflects an empirical judgment: in China’s system, advanced civilian platforms can be mobilized for military modernization.

The public record is imperfect but not empty. AP reported on June 8, 2026 that the Pentagon added Alibaba, BYD and Baidu, that DoD cited MIIT affiliation for them, and that the list is aimed not only at PLA-controlled firms but also companies contributing to China’s defense industrial base. The statutory text makes MIIT affiliation directly relevant, not incidental. For Alibaba, there are additional public allegations: a 2023 Senate letter states that Alibaba Cloud had MCF-related cooperation with Norinco, a major defense conglomerate, and with the National University of Defense Technology, and participated in a PLA Academy of Military Sciences military big-data forum. That is precisely the cloud-and-data nexus for which an early-warning designation is rational.

For Baidu and BYD, the public evidence is less specific, but the risk logic remains evidence-based rather than purely speculative. CSET’s dataset of 2,857 PLA AI-related award notices from 2023-2024 found 1,560 entities winning at least one AI-related contract, covering language and vision models, autonomous vehicles, data processing, logistics systems, smart manufacturing and robotics. That shows the PLA’s demand signal maps directly onto Baidu’s AI/autonomy stack and BYD-adjacent vehicle, battery, robotics and manufacturing ecosystems. The consequence is also real: Crowell reports that direct and indirect DoD procurement prohibitions tied to 1260H take effect on June 30, 2026 and June 30, 2027, and screening providers already use the list. A transparent rebuttal process should exist, but precaution is justified where dual-use platforms can become military infrastructure before a weapons contract appears.