9 February 2026 · 8 min
70% of Your API Comes from Overseas, and Washington Just Noticed
More than 70% of API manufacturing facilities registered with the FDA sit outside U.S. borders, predominantly in China and India. 90% of U.S. biotech companies rely on imported components for FDA-approved products. The BIOSECURE Act, the FDA's new PreCheck pilot programme, and escalating tariff threats are forcing a strategic rethink. But reshoring is a formulation problem as much as a procurement problem. When your API source changes country, your formulation changes with it. Different synthesis routes produce different polymorphs, different particle size distributions, different impurity profiles.
Where U.S. Drugs Are Actually Made
78% of FDA-registered API manufacturing sites are located outside the United States. Measured by production volume rather than facility count, the concentration worsens: approximately 80% of the APIs used in U.S. drugs come from facilities in China and India.
The U.S. Pharmacopeia's Medicine Supply Map, the most authoritative data source on API manufacturing concentration, lays out the current distribution. The United States accounts for just 22% of FDA-registered API manufacturing sites, down from higher historical levels. India holds 21%. China holds 20%, with 467 facilities as of February 2025, up from 230 facilities and a 13% market share in 2019. The European Union rounds out the picture at 19%.
The facility count understates the actual dependency. China and India's combined 40% share of registered facilities translates to roughly 80% of production volume, reflecting the scale and specialisation of Asian manufacturing operations relative to Western facilities. The pipeline is tilting further: in 2024, for the first time in more than two decades, China surpassed India in annual Drug Master File filings, capturing 45% of new API DMFs. China filed 392 API DMFs in 2024, nearly tripling from 134 in 2021. The U.S. share of the manufacturing pipeline is shrinking in real time.
Foreign API Sites
Of FDA-registered API manufacturing sites are outside the U.S.
Down from higher historical levelsChina's DMF Share
Of new API Drug Master File filings in 2024
Up from 13% in 2019Single-Source KSMs
Of key starting materials from a single country
Critical Concentration RiskThree Policy Actions in Twelve Months
The political establishment has noticed. Between May 2025 and February 2026, Washington launched three distinct policy mechanisms aimed at reshoring pharmaceutical manufacturing. Each addresses a different dimension of the problem, but none addresses the technical reality of what reshoring actually requires.
The BIOSECURE Act
Signed 18 December 2025, as Section 851 of the NDAA FY2026. Restricts U.S. federal procurement and grants involving biotechnology products or services from designated "biotechnology companies of concern." Prohibitions take effect approximately 970 days after enactment.
Regulatory Template for Decoupling
Tariffs & Executive Orders
Tariffs reaching up to 250% on pharmaceutical imports floated by August 2025, followed by a 100% tariff announcement on branded products effective 1 October 2025, unless manufacturers were building plants in America. Bilateral deals followed, including Pfizer's three-year tariff exemption.
Incentivises Announcements Over Capacity
FDA PreCheck Pilot
Launched 1 February 2026. Pre-qualifies domestic manufacturing facilities before they have products to submit. Seven participants will be selected for the initial cohort by 30 June 2026. A fundamental shift in FDA's posture from reactive to proactive, reducing one of the largest barriers to reshoring: uncertainty around whether a new facility will meet FDA approval standards.
Most Technically Significant Development
$480 Billion in Reshoring Pledges
The industry's response has been swift and enormous in dollar terms. Since Eli Lilly's announcement in late February 2025, fourteen major pharmaceutical companies have pledged a combined total of more than $480 billion over the next four to ten years in U.S. manufacturing investment.
The commitments include Merck at $70 billion, Pfizer at $70 billion (secured with a three-year tariff exemption), Johnson & Johnson at $55 billion, Roche at $50 billion over five years, Eli Lilly at $50 billion (including four new production facilities), AstraZeneca at $50 billion, Gilead Sciences at $32 billion, GSK at $30 billion, Novartis at $23 billion, and Novo Nordisk at $10 billion additional, including end-to-end U.S. production for Wegovy tablets.
The announcement headlines obscure a timing problem: industry leaders acknowledge that groundbreaking on most U.S. pharmaceutical investments will not happen before 2026 or 2027, with validated operational capacity following years after that. Capital can fund construction, but the workforce, quality systems, and validated supply chains take longer to build than the buildings themselves.
| Company | Pledge Amount | Timeline | Notes |
|---|---|---|---|
| Merck | $70B | 4-10 years | |
| Pfizer | $70B | 3-year tariff exemption | Bilateral deal |
| J&J | $55B | Not stated | |
| Roche | $50B | 5 years | |
| Eli Lilly | $50B | Not stated | Four new facilities |
| AstraZeneca | $50B | Not stated | |
| Others | ~$135B | Various | Gilead, GSK, Novartis, Novo Nordisk |
| Total | $480B+ | 2029+ earliest validated production |
Every Source Change Is a Reformulation Event
The gap in the policy conversation is technical, not logistical. Washington's response (tariffs, reshoring mandates, BIOSECURE restrictions, FDA PreCheck) treats pharmaceutical dependency as a procurement problem. Move the sourcing, solve the problem. But every pharmaceutical formulation scientist knows that when you change your API source, you change your formulation.
An API manufactured in Hyderabad is not the same material as the same molecule manufactured in North Carolina. Synthesis routes, crystallisation conditions, and milling processes all vary between facilities, producing material that may be chemically identical but physically distinct.
The regulatory burden is substantial. Under FDA's SUPAC and BACPAC guidance frameworks, any change in the site of API manufacture is classified as a major change, requiring a Prior Approval Supplement. Companies must demonstrate no adverse impact on quality, purity, stability , or performance. A BIO survey from February 2025 found that 80% of biotech firms need at least 12 months to find alternative suppliers, and 44% would require more than two years.
Polymorphic Form
Different crystalline arrangements of the same molecule dissolve at different rates. Abbott's ritonavir proved this when a new polymorph (Form II) appeared, less than 50% as soluble as Form I, forcing a complete reformulation and market withdrawal.
Particle Size Distribution
Different API manufacturers use different crystallisation, milling, and micronisation techniques. A source change can shift particle size from D50 of 15 microns to D50 of 45 microns, completely altering the dissolution profile and bioavailability.
Impurity Profile
Different synthetic routes to the same API molecule produce different impurities. A new source may introduce impurities that interact with excipients, catalyse degradation, or fall outside established ICH Q3A/B limits, requiring new analytical methods and stability evaluation.
Surface Properties
Wettability, powder flow, and morphology (crystal shape: needles, plates, prisms) all affect processability. Material that flows smoothly through a tableting press from one source may cause weight variation and content uniformity failures from another.
U.S. Drug Shortages Are Already at Record Levels
The United States has a persistent drug shortage problem that makes supply chain disruptions immediately dangerous. According to the USP's 2024 Annual Drug Shortages Report, 89% of drug shortages experienced in 2024 carried over from 2023. Long-standing, persistent shortages account for more than 90% of all drug shortages in the United States.
The average duration of current shortages has risen to more than four years. More than 40 medicines have been in shortage for over three years. Five medicines (atropine sulfate, fentanyl citrate, leucovorin calcium, lidocaine hydrochloride, and epinephrine bitartrate) have been in shortage for 10 or more years. All five are used in paediatric care, including anaesthesia, pain management, and oncology.
The ASPE/HHS analysis of shortages between 2018 and 2023 found that 258 unique active ingredients went into national shortage, represented by 1,961 unique CDER-regulated prescription drug products. Injectable drug products made up 50% of all shortages, with 65% of injectable shortages still unresolved after three years. The median duration of an injectable shortage is 4.6 years.
Excipient Supply Chains Are Equally Exposed
The policy conversation focuses almost exclusively on APIs, but excipient supply chains are equally vulnerable and less closely monitored. Major pharmaceutical excipients with significant Chinese manufacturing concentration include hydroxypropyl methylcellulose (HPMC), magnesium stearate, microcrystalline cellulose, silicon dioxide, povidone, and lactose. Excipient supply chains involve multiple intermediaries, and traceability often breaks down.
The consequences of excipient supply chain failure are not hypothetical. In March 2008, contaminated raw heparin stock imported from China caused the death of 81 people and 785 reports of serious injuries. The contaminant was oversulfated chondroitin sulfate, not a naturally occurring molecule, costing a fraction of true heparin. The contamination was almost certainly intentional.
Excipient source changes carry their own reformulation burden. Different excipient sources can vary in particle size and morphology, moisture content, compactability, functionality-related characteristics, and trace impurity levels. These variations can alter formulation performance, requiring compatibility testing, stability studies, and potentially updated regulatory filings. For companies reshoring or diversifying both API and excipient sources simultaneously, the reformulation workload compounds.
Building Formulations That Survive Source Changes
The ICH Q8(R2) framework for Quality by Design offers the conceptual solution: design spaces broad enough to accommodate source variation. A well-defined design space establishes the multidimensional combination of input variables (critical material attributes and critical process parameters) that have been proven through experimentation to assure quality. If API particle size from Supplier A is D50 of 15 microns and from Supplier B is D50 of 40 microns, a formulation and process built for that range should produce the same critical quality attributes across it.
In practice, most pharmaceutical companies have not built formulations with this degree of source flexibility. Development programmes typically optimise around a single API source and a narrow set of excipient suppliers. When that source changes, the formulation breaks.
Multiply this by dozens or hundreds of products in a portfolio. Adding headcount to a formulation development team will not solve it. The combinatorial complexity of material attributes, process parameters, and quality targets across multiple products and multiple potential sources exceeds the capacity of traditional empirical development.
How DeepC Addresses This
The reshoring transition creates four distinct technical challenges that converge on formulation. DeepC's studies take on each of them.
Source-Robust Formulation Design
DeepC's formulation study opens with a QTPP that sets out the API's critical material attributes, then designs candidates with an ICH Q9 risk assessment and a control strategy, so the attributes a new source could shift, such as polymorphic form, particle size, and impurity profile, are named before qualification starts.
Adjusting to a New Source
Give the formulation optimization an existing formulation and the new API's attributes, and it returns up to three variants, each with a before-and-after table, its trade-offs, and the next experiments to run. DeepC Optimize, in early access, then proposes each round of bench work from your results.
Excipient Substitution Intelligence
DeepC's literature review identifies alternative excipients that match the function of current ones, with IID precedent for the route and dosage form, commercial grades from PharmaExcipients, and the Drug Master File holders behind each. ANVISA's GMP certificates show which sites are authorised to make them, where, and until when.
Portfolio-Level Source Risk
Ask product by product which ingredients rest on a single DMF holder or a single certified site, with every answer cited to its register. Bring your own portfolio data to the data analysis to rank where reformulation investment should go first.
BIOSECURE, tariff threats, FDA PreCheck, and the ongoing drug shortage crisis together point to six concrete actions:
- The 2026-2028 window before BIOSECURE restrictions take full effect is the period of greatest exposure. Companies that have not mapped their KSM-to-finished-product supply chains at a granular level lack the visibility to act on source diversification.
- Every API source change triggers reformulation, analytical, stability, and regulatory work. Source diversification is a formulation programme, not a procurement exercise. Companies that staff and budget accordingly will execute faster.
- For new products entering development now, building formulations with wide material attribute acceptance ranges is insurance against sourcing requirements that can shift overnight. ICH Q8(R2) provides the framework; AI-powered formulation platforms provide the computational power to define design spaces that would take years to map experimentally.
- Injectable drugs account for 50% of all shortages, with a median shortage duration of 4.6 years. These products are the most likely to face supply disruption during the reshoring transition and the ones where reformulation failure has the most direct patient impact.
- 80% of companies need 12+ months just to identify alternative suppliers, before reformulation begins. A facility breaking ground in 2026 will not produce validated, FDA-approved product until 2029 or 2030 at the earliest. Reshoring timelines are longer than the announcements suggest.
- The seven slots in the initial PreCheck cohort reduce the largest source of uncertainty in domestic facility investment: whether the facility will meet FDA approval standards. Companies with strong formulation data packages will be best positioned to demonstrate manufacturing readiness.
$480B
in reshoring pledges. Zero invested in formulation robustness. Washington can mandate reshoring, impose tariffs, and pre-qualify facilities. None of these policy mechanisms changes the physical reality: an API manufactured in a different country is a different material. Formulation robustness requires science, and increasingly, AI.

