Semaglutide's Patent Cliff: A Global Fracture, Not a Single Event

The Date That Already Passed

On March 20, 2026, the twenty-year clock ran out on Novo Nordisk’s foundational semaglutide patent — the compound patent, filed as international application WO2006097537A2 on March 20, 2006, covering the molecule itself rather than any particular formulation, dose, or delivery device. In every jurisdiction that applies a standard twenty-year patent term without extension, compound protection for the active ingredient behind Ozempic, Wegovy, and Rybelsus simply ended that day.

That sentence undersells how strange the actual outcome was. A single patent, filed on a single date, expiring on a single twentieth anniversary, nonetheless produced a genuinely different competitive reality in different countries within the same week — because patent term extensions, supplementary protection mechanisms, and national patent law simply do not treat a twenty-year-old filing the same way everywhere. Understanding why is most of what matters here, more than the headline date itself.

Why One Molecule Has Four Different Expiration Dates

In India, China, Brazil, and South Africa — jurisdictions applying the standard twenty-year term without the kind of extension mechanisms available elsewhere — compound protection for semaglutide ended on schedule, March 20, 2026. In Canada, the situation was even more abrupt: the granted patent in the same family, CA2601784C, was allowed to lapse outright, with no legal remedy available to restore it.

Europe tells a different story. Semaglutide there benefits from Supplementary Protection Certificates, a mechanism designed to compensate originators for the years a drug spends in regulatory review before it can be sold — time that would otherwise eat into the effective commercial life of a twenty-year patent. Those certificates push European protection out to March 2031. In the United States, patent term extensions and adjustments tied to a specific patent, US 8,129,343, are expected to hold core protection until December 2031, with additional formulation and device patents tied to Ozempic specifically extending as far as 2038. Japan sits close to the European timeline, with protection expected to run to roughly 2031.

None of this is exotic or improper — patent term extension for regulatory delay is a standard, well-established feature of pharmaceutical patent law in most developed markets, not a Novo-specific maneuver. But the practical effect is that a molecule which, on paper, lost its foundational patent protection on a single day in March 2026 will not face genuine, unrestricted generic competition in its largest and most profitable markets for another five to twelve years past that date, depending on jurisdiction.

The India and China Response

In the markets where protection genuinely ended on schedule, the response was immediate and large. India saw rapid generic entry and visible price cuts within weeks of the March expiration, consistent with how the country’s generics industry has historically moved on major patent cliffs — India’s combination of large-scale peptide manufacturing capacity, built out over the preceding several years partly in anticipation of exactly this event, and a domestic regulatory apparatus well practiced at fast-following a loss of exclusivity, meant generic semaglutide reached pharmacy shelves at a pace that would be difficult to replicate in a market with less manufacturing depth already in place.

China’s response has been differently shaped. At least fifteen Chinese pharmaceutical companies were already in generic development ahead of the expiration, positioning the country to become plausibly the largest single semaglutide market once generic pricing takes hold, given its patient population. But a notable domestic legal development complicates the simple “generics flood in” narrative: reporting from late 2025 indicates a court upheld a Novo secondary patent — not the core compound patent, which expired on schedule, but a follow-on patent relying in part on supplementary experimental data — a ruling read by legal observers as likely to push domestic competition toward genuine next-generation reformulation and improvement rather than pure molecule-for-molecule copying. If that reading holds, it reinforces a broader shift already underway in Chinese pharma, away from fast-follower generics and toward first-in-class ambition, discussed elsewhere in this issue’s coverage of the AI-native drug discovery boom.

The Patent Thicket Behind the Headline Patent

It would be a mistake to treat the compound patent as the whole story, because Novo has not been sitting still for the twenty years since that original filing. Across Ozempic, Wegovy, and Rybelsus combined, the company has filed roughly 320 US patent applications and been granted 154 patents — covering injection pen mechanisms, specific formulations, dosing regimens, and methods of treatment, layered on top of the core molecule patent that just expired in most of the world.

That density of secondary patent coverage is what patent lawyers call a thicket, and it is a standard, if aggressive, feature of how large-molecule blockbusters defend themselves once the foundational patent clock starts running out — the same defensive architecture examined in detail elsewhere in this issue with respect to Keytruda’s own 2028 cliff. A generic manufacturer that clears the compound patent hurdle in a given market still has to navigate whichever of those 154 granted patents happen to be enforceable in that specific jurisdiction, which is precisely why the practical, on-the-ground generic entry date varies so much more than the simple "molecule patent expired" headline would suggest.

The Compounding Wildcard, Which Has Nothing to Do With Any of This

One further complication is worth separating out clearly, because it gets conflated with the patent story constantly and the two are legally unrelated. Compounded semaglutide has been available in the United States since 2022, not because any patent expired, but because semaglutide sat continuously on the FDA’s drug shortage list, which triggers a separate statutory exemption under Section 503A of the Food, Drug, and Cosmetic Act allowing licensed compounding pharmacies to prepare copies of an FDA-approved drug during a genuine supply shortage.

Patent law and drug-shortage policy are entirely separate legal frameworks, and it is worth being precise about that distinction: patent expiration does not automatically make compounding legal, and shortage-driven compounding exemptions make copies legal even while the underlying patents remain fully enforceable. As supply constraints ease — and Novo’s manufacturing capacity has been expanding specifically to address the shortage that has persisted since March 2022 — the compounding pathway is likely to close well before the underlying US patents actually expire, which will mean a genuine gap between when unauthorized compounded semaglutide disappears from the US market and when authorized generic competition eventually arrives around 2031 to 2032.

A Third of the World’s Obese Adults, In Round Numbers

It is worth sizing the population actually affected by the markets where protection has already lapsed. Life sciences analytics firm IQVIA has estimated that the countries where Novo’s patent has already expired — India, China, Brazil, and South Africa chief among them — represent roughly a third of the global adult population living with obesity. That is a genuinely large fraction of the addressable patient population gaining access to lower-cost semaglutide years before the US and European markets will see the same thing, and it is a meaningfully different distribution of benefit than most prior blockbuster patent cliffs, which have tended to hit the US and Europe first and emerging markets only years later through voluntary licensing or parallel importation.

Where the Investable Value Sits

For a venture investor, the direct opportunity in semaglutide generics themselves is largely a manufacturing-scale and regulatory-execution play rather than a venture-stage bet — the companies positioned to win in India and China were, for the most part, already large, established generics manufacturers with the bioreactor and peptide-synthesis capacity to move fast, a dynamic covered at greater length elsewhere in this issue’s look at India’s CDMO sector, which has been building exactly this kind of GLP-1-specific capacity for several years in direct anticipation of this event.

The more interesting venture-stage angle sits one layer removed: companies building the analytical comparability testing, formulation science, and device-patent workaround engineering that a second-wave or third-wave generic entrant needs to clear Novo’s secondary patent thicket in the markets where the compound patent alone is not sufficient protection to unlock the market — the US and Europe specifically, where genuine generic entry is still several years out and where the eventual winners will be the companies that solved the formulation and device patents years in advance rather than waiting for 2031 to start.

The Pattern Worth Remembering

The lesson from semaglutide’s cliff that will matter most for the rest of this decade is a structural one, not a Novo-specific one: a single molecule patent expiring on a single date does not produce a single market event. It produces a staggered, jurisdiction-by-jurisdiction unlocking that can span the better part of a decade between the first market where competition arrives and the last, shaped as much by patent term extension mechanics and secondary patent thickets as by the underlying twenty-year clock everyone quotes as the headline number. Tirzepatide, examined separately elsewhere in this issue, is now heading toward its own version of exactly this pattern — on a different timeline, and, as it turns out, a rather different structure.