Pick up a packet of corn, soy, or tomato seed off the shelf at any large agricultural retailer, and there's better than even odds it came out of a lab owned by one of four companies. Not dozens. Not hundreds. Four. As recently as the mid-2010s there were six, each with its own history, its own patent portfolio, its own slice of the market. Three years of mergers turned six into four. And those four now hold something no harvest on the planet can happen without: the right to decide what the next generation of a plant will be, before it has even sprouted.
Look at the list. It's shorter than it sounds.
Who holds it
Between 2017 and 2018, the agrochemical and seed industry went through a wave of consolidation that redrew the map of world agriculture in three years. Bayer bought Monsanto for roughly $63 billion — a deal that put the world's largest herbicide maker and the world's largest holder of GMO seed patents under one roof. Dow and DuPont merged into DowDuPont, then split into three companies, one of which, Corteva, inherited the combined seed and agrochemical business of both parents. China's state-owned ChemChina acquired Switzerland's Syngenta for around $43 billion, gaining access to seeds, pesticides, and — more importantly for Beijing — a stake in global food security infrastructure. BASF picked up a chunk of the assets Bayer was forced to divest to get antitrust approval for the Monsanto deal in the first place.
The result: Bayer, Corteva, ChemChina (which now owns Syngenta), and BASF — by various estimates, this group controls somewhere around half, and by some counts more, of the world's market for proprietary, patent-protected seed. Add a comparable share of the agrochemical market, since these companies sell seed and the chemicals it's designed to be treated with as a bundle, not as separate goods. ETC Group, which has tracked concentration in agribusiness for years, doesn't call this a market. It calls it an oligopoly — a handful of players whose pricing, variety-availability, and licensing decisions touch every eater on the planet, whether they've ever heard the company names or not.
How a seed became property
Worth pausing here, because the fact that a seed can be patented at all is not a law of nature. It's a legal construct barely forty years old.
Before 1980, a living organism could not be patented in the United States as a matter of principle — the position was that living things belong to nature, not to an inventor. The Supreme Court's decision in Diamond v. Chakrabarty opened the door: a genetically modified organism was ruled to be a patentable invention. From there, the logic ran downhill. Companies stopped patenting whole varieties — that had long been possible under a separate plant-breeder's-rights regime — and started patenting the specific engineered traits inside them: herbicide resistance, pest resistance. Buy a seed carrying that trait, and you sign a licensing agreement that explicitly bars you from saving harvested seed for next year's planting. Want to plant again, buy again.
That upended a practice as old as farming itself. From the domestication of crops through the end of the twentieth century, a farmer setting aside part of the harvest as seed for next season wasn't a gesture — it was how agriculture mechanically worked. Bowman v. Monsanto, decided by the U.S. Supreme Court in 2013, fixed the new norm in place: a farmer who planted a second generation of patented soybeans without a fresh licensing agreement infringed the patent, even though he'd bought that seed legally, off a grain elevator, as ordinary commodity grain. The Court ruled against him unanimously. From that point it was legally settled: the right to a seed's offspring belongs not to whoever planted it, but to whoever holds the patent.
The price of one gene
What that means in practice, outside the courtroom.
The price of patent-protected seed carrying an engineered trait has outpaced inflation for years, because a farmer isn't just paying for biology — they're paying for a license on the gene, and often for a matched chemical package the gene is designed to work alongside. Herbicide resistance is only useful paired with that same manufacturer's herbicide; seed and chemical are sold as one system, not two products you can mix and match. A farmer who enters that system once has a hard time leaving it — unpatented alternative varieties are often less productive in the short run, simply because most of the last several decades of breeding effort has concentrated on a handful of commercial lines rather than on breadth.
Which brings up a second, quieter cost — not measured in money, but in diversity. By the UN Food and Agriculture Organization's estimate, humanity has lost on the order of three-quarters of the genetic diversity in food crops that was once cultivated in fields, over the course of the twentieth century. Thousands of local wheat, corn, and rice varieties, each selected over generations for the soil and climate of one particular valley, have gone out of circulation as the market consolidated around a small number of high-yield, patented lines sold identically everywhere. That diversity was never a museum piece. It's the species' insurance policy against a future disease, drought, or pest that today's dominant varieties may not be built to survive. A harvest that's genetically uniform across millions of acres at once is one good year away from one bad pathogen mutation away from catastrophe.
Our record: Ren — the name that grants power to whoever speaks it first — applies literally here. To name a gene, to register it as an invention, is to claim the right to decide its fate for decades. A seed carries a plant's Ka, its life-force, capable of replicating itself without asking permission. A patent is an attempt to install a toll on exactly that moment of self-replication — the one point in the food cycle nature has historically given away for free.
Where the seed still belongs to many
No concentration is ever total, and there are counterexamples worth knowing.
The Global Seed Vault, built by Norway into permafrost on the island of Svalbard for exactly this kind of catastrophe scenario, is owned by no agribusiness. It's a backup, not a bank: national and international gene banks around the world deposit samples there, each depositor keeps ownership of its own seed, and the vault simply holds a duplicate for the worst case. In parallel, since 2014, the Open Source Seed Initiative in the U.S. and similar efforts elsewhere have breeders deliberately releasing new varieties under a license that legally bars anyone from patenting their descendants in the future, keeping that genetics in open circulation permanently. Community seed banks — in India, across sub-Saharan Africa, in farming cooperatives worldwide — carry on the practice of selecting and exchanging varieties the way it was done before the concept of a gene patent existed at all.
None of this operates at the scale of four agribusiness giants. But it's proof the alternative isn't hypothetical — it already exists, already works, and needs no regulator's permission to keep going.
The seed is the smallest object in this conversation and the most fundamental one at once: the little that holds an entire next life folded inside itself, waiting on water. Whoever decides what grows out of that particular packet is deciding the shape of next year's table for billions of people who will never learn their name. That's worth holding not as a reason for alarm, but as a fact you can check — starting with the label on the next seed packet that ends up in your hands.