In October 2025, the lights went off in Union, South Carolina. MycoWorks, the company that had convinced Hermès to wrap one of its handbags in mushroom skin, ran out of money. Its 136,000-square-foot factory, opened with great ceremony two years earlier and built to grow several million square feet of fungal leather a year, had been running at roughly a fifth of capacity. The company filed for an assignment for the benefit of creditors, the quiet cousin of bankruptcy, and shut the plant down. WWD called it one of the most-watched biomaterials companies in the world. By Halloween it was being liquidated.

Two years before that, Bolt Threads had already pulled the plug on Mylo, the mushroom leather that Adidas, Kering, Lululemon, and Stella McCartney had all queued up to use. Co-founder Dan Widmaier told Vogue Business the material had come “devastatingly close” to commercial scale before the funding dried up. “Scale-up is the most expensive and hardest part,” he said, “and there are no shortcuts.”

So the obvious headline writes itself: mushroom leather was a bubble, and the bubble popped.

Except the patent record says the opposite. US grants that mention mycelium hit 21 in 2025, the highest annual total on record, up from 16 in each of the prior two years and roughly double the count of a decade ago. The all-time peak arrived in the same calendar year the field’s flagship startup went insolvent. Something is filing these patents. It just isn’t the startups anymore.

What the fungus actually does

First, the mechanism, because it is the part that survives every funding cycle. Mycelium is the root structure of a fungus: a dense, self-branching web of filaments called hyphae. Feed it a tray of agricultural waste like sawdust or hemp hurd, control the air, and it will knit that loose substrate into a continuous solid mat. You are not manufacturing a material. You are growing one, and it assembles itself for the price of the feedstock and a few weeks of time.

The trick that separates a science-fair demo from a product is control, and that is exactly what the recent patents claim. Ecovative, the Rensselaer Polytechnic spinout that two students started in 2007, holds a grant on growing “mycological biopolymers in void space tooling.” Read the claims and the engineering is precise: you pack a mold with substrate and fungus, seal it except for one outlet, and hold the chamber at 5 to 7 percent carbon dioxide and 85 to 95 degrees Fahrenheit. Those conditions are not arbitrary. High CO2 keeps the fungus from doing what fungi want to do, which is sprout a mushroom. No cap, no stem, no spores. The organism stays in its undifferentiated state and simply fills the mold with pure chitinous mat, taking the shape of whatever void you gave it. It is injection molding where the resin is alive.

MycoWorks, before it folded, solved a different problem: how to peel a clean sheet off the tray. Its patents describe a “mycelium growth bed” with a perforation layer buried between the substrate and the growing material. The fungus grows through the perforations, which engineer a deliberate, uniform line of weakness so the finished sheet delaminates in one even piece instead of tearing raggedly. That sounds trivial. It is the difference between a lab curiosity and something a tannery can run by the roll.

Mentally delete the word “mushroom” from all of this and the inventions still belong together: a self-assembling biopolymer, grown to net shape, harvested as a uniform sheet, tuned by gas and temperature. That is a manufacturing platform, not a buzzword.

The brands are insourcing the biology

Here is the part nobody reported. In April 2025, five months before MycoWorks went under, Hermès was granted its own US patent. Not a partnership announcement, not a marketing collaboration: a method patent, assigned to Hermès Sellier, for “producing a finished leather substitute.” The claimed invention is narrow and revealing. You take a sheet grown from mycelium, identify which face grew first, and apply a finishing coating to the bottom surface only. The patent lists the end products by name: a bag, shoes, a watch strap, a belt, a wallet.

Hermès had been buying its mushroom leather from MycoWorks since 2021, when it showed a version of its Victoria travel bag rendered in a MycoWorks material it branded Sylvania. A luxury house that buys a material from a supplier does not normally file patents on how to finish it. A luxury house that has watched its supplier wobble, and that intends to keep selling the product after the supplier is gone, might.

Then there is Nike. On a single day in August 2025, the company was granted two patents that the trade press missed entirely. One covers a “biocomposite material” with “zonal properties,” made by growing mycelium from a fungal strain around and through an internal scaffold. The second applies it to a “wearable article.” Buried in the claims is the specific thing: an article of footwear with an upper internal structure that has a lacing zone, a toe zone, and side zones, where the structure is permeable to mycelium growth in some areas and not others. Nike is describing a shoe upper where the fungus is grown directly into a frame, stiffer where it needs to grip a lace, softer where it wraps the foot. The material and the geometry are produced in the same step.

This is the adjacent-possible move in its purest form. The biology was proven by startups that then ran out of runway. The brands with the balance sheets, the factories, and a reason to own leather are now picking up the patents the pure-plays could not afford to scale. Bolt Threads, even after pausing Mylo, was granted a continuation on its core composite-material patent in February 2024, the IP outliving the product line by a year.

Who survives, and where it spreads

Not everyone is dead. Ecovative, the oldest player, is still standing, with around 15 US mycelium grants to its name across its corporate variants and a stated plan to launch commercial leather and foam products with brand partners in 2027. Its packaging business, the mushroom blocks that once shipped inside Dell computers and that IKEA committed to using in place of polystyrene, gave it a revenue floor the leather-only startups never had. The lesson of the past three years is brutal and simple: the companies that tried to sell only grown leather are gone, and the ones that sold grown leather plus something boring enough to pay the bills are not.

The mechanism, meanwhile, keeps wandering into rooms nobody expected. One Ecovative patent describes mineralizing a mycelium scaffold with silicates and apatites, proposing it as a stand-in for the collagen and hydroxyapatite membranes used in medical bone repair. Stranger still, Boeing holds two grants on using a mycelium structure, with a surface area above 800 square meters per cubic meter, as a sponge for storing hydrogen inside a pressure vessel. The same self-branching filament web that makes a handbag feel like calfskin also makes an enormous amount of surface in a small volume, and surface area is exactly what a gas-storage engineer is shopping for.

That is the signal worth watching. Press coverage tracks startups, and startups die loudly. Patents track capability, and capability is quietly migrating from companies whose names you knew to companies whose names you know better. In April 2026, designers in Milan showed a shoe with a load-bearing sole grown entirely from mycelium, no startup attached. The obituaries were premature. The fungus is still growing. It just changed landlords.

Method note. Counts come from 9.3M US utility patent grants sourced from USPTO bulk grant data, searched for full-text mentions of mycelium and grouped by grant year and by assignee, with variant company spellings and subsidiaries combined. The materials-focused subset was identified by reading the abstracts and claims of grants held by MycoWorks, Ecovative, Hermès Sellier, Nike, Bolt Threads, and Boeing to confirm they describe growing structural or sheet material from fungal hyphae, as distinct from the separate cluster of patents on fungal protein for food. Annual mention counts include all uses of the term, so the totals are a directional indicator of activity, not a clean materials-only census. Company history, funding, and shutdown details are drawn from reporting in WWD, Business of Fashion, Vogue Business, and phys.org, and from public company records. Grant dates and claim language were read directly from the patent documents.