Dragon Bio — May 08, 2026

This isn't the week the US-China biotech competition turned — it's the week the clocks all became visible at once. The BIOSECURE Act is the law of the land, but the OMB list that actually names companies isn't due until December 2026. Samsung Biologics keeps building capacity that US CDMOs can't match. China keeps approving CRISPR crops faster than the USDA can process the paperwork. The signals are administrative, not dramatic. Which is exactly how strategic competition usually looks before it becomes obvious.

The National Security Commission on Emerging Biotechnology, chaired by Senator Todd Young (R-Ind.), published its capstone report on December 19, 2025, arguing the United States has roughly 36 months to mount a coherent response before Chinese biotech leadership becomes structurally entrenched. The commission recommended at least $15 billion in federal investment over five years and a new National Biotechnology Coordination Office inside the White House to pull together what is currently a scattered US response across USDA, HHS, DoD, Commerce, and the intelligence community.

The substantive claim is that China has built a vertically integrated biotech ecosystem operating on non-market logic — capacity, market share, and strategic positioning ahead of margin — and that WuXi AppTec and WuXi Biologics are the most visible examples. What changes if Washington takes this seriously: a coordination office with budget authority would let CFIUS, BIS export controls, BARDA's manufacturing build-out, and USDA approvals stop working at cross-purposes. What failure looks like: the report becomes a citation in next year's report. The signal to watch is whether the FY2027 budget request, due early next year, contains anything resembling the $15 billion figure or whether it gets whittled to a coordinator title and a website.

President Trump signed the BIOSECURE Act into law on December 18, 2025, as part of the FY2026 NDAA, restricting US federal procurement and grants involving designated "biotechnology companies of concern." But the statute itself doesn't name companies — that's OMB's job, with a deadline of December 2026. According to legal analysis from Latham & Watkins and Morrison Foerster, full enforcement against 1260H-listed entities won't be effective for up to 970 days after enactment. So the law is real, the timeline is glacial, and the lobbying war is just beginning.

What changes if WuXi AppTec, WuXi Biologics, and WuXi XDC end up on the December list: the five-year grandfather clock starts, US pharma transitions move from "phased" to "scheduled," and Korean and European CDMOs become the default replacement. What failure looks like: OMB designates a small number of marginal entities, leaves WuXi off, and the law becomes a paper tiger that complicates compliance without changing supply chains. The signal: whether DoD's next 1260H update — overdue on its annual cadence — adds the WuXi entities. As of this writing, no addition has been confirmed.

Samsung Biologics announced completion of its acquisition of GSK's Rockville, Maryland facility, adding two cGMP plants and roughly 60,000 liters of drug-substance capacity to its US footprint. Combined with its Songdo expansion, Samsung's total antibody manufacturing capacity is on track to exceed 600,000 liters — larger than the combined capacity of US-headquartered CDMOs. Samsung's Q1 2026 results, released April 23, document continued backlog growth and the operational handoff at Rockville.

What changes if US pharma routes BIOSECURE-driven transitions to Samsung: the Act's intent — reducing dependence on non-allied biomanufacturing — gets quietly inverted. Korea is an ally, but Korea is not the United States, and the domestic CDMO capacity Washington keeps saying it wants doesn't get built. What failure looks like for Samsung: backlog growth flattens as Lonza, Catalent, and Fujifilm Diosynth expand, or BARDA's Contract Development Institute awards finally produce credible domestic alternatives. The signal: Samsung's Q2 contract announcements and whether any of them name a former WuXi client.

China's Ministry of Agriculture and Rural Affairs (MARA) issued a new batch of safety certificates for gene-edited crops on May 5, 2026, including a blast-resistant rice variety developed by the Chinese Academy of Agricultural Sciences. The Genetic Literacy Project's tracker continues to document a sustained Chinese lead in CRISPR crop approvals — though as CEPA noted in its analysis, MARA reduced the public visibility of biosafety certificate issuance in 2025, making real-time Western monitoring meaningfully harder.

What changes if China deploys drought-tolerant, disease-resistant, yield-improved varieties at commercial scale five years before the United States: Chinese agricultural productivity compounds, Chinese seed varieties propagate through Belt and Road agricultural partnerships, and US export competitiveness in commodity grains erodes in the markets where the productivity gap matters most. What failure looks like: field performance disappoints, adoption stalls at the provincial level, and Chinese approvals turn out to be regulatory theater rather than deployment. The signal to watch is whether the USDA accelerates its own CRISPR approval cadence in response or continues processing roughly two to three approvals per year.

DARPA's Biological Technologies Office published a January 2026 program solicitation seeking teams to build an optically-controlled, template-free system for synthesizing nucleic acid sequences inside living cells — a "nucleic acid compiler" that would let a user program new genetic instructions into a cell using nothing but light. The Gate 1 abstract deadline has passed; the program is now in oral presentation review.

The framing of the solicitation is the actual signal. DARPA is telling potential performers, in the open, that current DNA synthesis and delivery methods are too slow, too dependent on external reagents, and too logistically vulnerable for the environments DoD cares about. What changes if the program produces a working prototype in its 33-month performance window: forward-deployed biomanufacturing stops needing cold-chain reagent supply, and the threshold for who can produce engineered biology drops sharply. What failure looks like: the program quietly extends or terminates, and the field returns to chasing better delivery vectors. The companion DARPA Protean program, with Gate 1 abstracts due March 12, 2026, and now in selection, is asking for protein-level chemical countermeasures effective at 10,000x LD50 — performance only achievable through computationally designed proteins. Both programs read partly as responses to published work from China's Academy of Military Medical Sciences.

South Korea's National Assembly passed the Synthetic Biology Promotion Act in April 2026, establishing a government-backed framework for Korean synthetic biology platforms, fermentation infrastructure, and bio-manufacturing capacity. The Ministry of Science and ICT followed with an initial $300 million tranche aimed at the K-BIO cluster across Songdo, Osong, and Daejeon, explicitly targeting the gap between academic R&D and commercial CDMO scale-up.

What changes if Korea executes: a third pole emerges in Asian biotech that isn't a US proxy and isn't a Chinese counterweight — it's a sovereign capability sitting between them, accepting business from both. Samsung's manufacturing dominance is one pillar; the new Act and the MSIT funding are another. Lotte Biologics' May 4, 2026 KRX filing disclosing a 120,000-liter expansion at Songdo, financed by an advance purchase agreement, is the third. What failure looks like: Korean policy ambition outruns execution, the funding gets diluted across too many institutes, and Korea ends up as a Samsung-and-everyone-else market. The signal: whether Korean synthetic biology startups start attracting US and European venture capital at scale, or remain a domestic-funded ecosystem.

As of November 28, 2025, supplying a defined novel food in Singapore without pre-market approval is an offense. The Singapore Food Agency's updated approved-novel-foods list, published March 2026, now includes precision-fermentation products like 3'-sialyllactose and 6'-sialyllactose produced in engineered E. coli — approved June 18, 2025 — alongside cultivated meat and Solein. The agency also issued separate public-facing guidance treating fermentation-derived foods as their own regulated category rather than fitting them into older food-additive categories.

What changes if more Asian governments adopt Singapore's framing: the first-mover regulatory geography for alternative proteins consolidates in Singapore, and the question of where to launch a precision fermentation product becomes settled rather than negotiated market by market. What failure looks like: Hong Kong, Tokyo, or Seoul build parallel frameworks that fragment the Asian regulatory map. The signal: which jurisdiction issues the next fermentation-specific framework, and whether it cites Singapore's definitions. The US still has more capital and more science. Singapore has the bureaucracy that turns frontier biology into a governed product category — which is what investors actually price.

The Pentagon wants to write DNA into cells using flashlights, Samsung is buying American factories one Maryland suburb at a time, and somewhere in Beijing a bureaucrat just finished cataloging 1.3 million patents while the USDA approved its fourth CRISPR crop of the biennium. The three-year warning window opened in December 2025; five months in, the loudest sound is paperwork. Until next week.

Forward this to the friend who keeps insisting biotech isn't a national security story.

• Defense (May 2, 2026): Trump declared the Iran war "terminated" on the 60-day War Powers deadline while leaving every asset in place. Read → Trump Declares the Iran War "Terminated"
• Critical Minerals (May 6, 2026): Japan, France, and Canada are quietly building a minerals bloc that doesn't put Washington at the center. Read → The G7 Is Quietly Building a Minerals Bloc
• Quantum (May 8, 2026): China unveiled a dual-core quantum computer that doesn't need liquid helium — a hardware divergence that rhymes with what's happening in biotech IP. Read → China Builds the World's First Dual-Core Quantum Computer