Denmark’s Monstrous Human Rights Abuses in Greenland
Copenhagen took Greenlanders’ children, violated their bodies, and forfeited any right to rule their distant colony. Greenlanders want independence, but can’t afford it. There’s a solution.
Amarok Petersen was 13 years old when Danish doctors implanted an IUD in her body. No one told her. She discovered it only when she was 27, after years of severe uterine problems, when another doctor finally found the device. By then, the choice to have children had been stolen from her. Years later, she learned that doctors had also removed her fallopian tubes without telling her.
“The Danes don’t see us as humans,” she said.
Last week, two expert investigations put the weight of an official record behind what Petersen and thousands of other Greenlandic women — out of a total population of just 57,000 — had tried to tell for years. Denmark’s treatment of Greenlanders was not “benevolent stewardship” marred by “a handful of mistakes.” It was colonial abuse inflicted on their bodies, their children, their families and their sovereignty.
It also conveniently produced exactly the result Copenhagen wanted: Greenlanders’ birthrate collapsed, the growth of the native population slowed sharply, the percentage of the population that’s Danish thus increased, and Denmark’s annual costs to maintain the colony were controlled.
The new data demolishes Copenhagen’s claim to be Greenland’s benevolent guardian. Denmark has forfeited the moral authority — and the moral right — to rule the island. Greenland’s future rightfully belongs to the Greenlanders. The problem is, they can’t afford the independence most of them want.
There’s a solution.
Their Bodies, But Not Their Choice
Denmark’s forced contraception campaign was staggering in scale. By the end of 1970 alone, nearly half of all women of childbearing age or younger had been fitted with IUDs, mostly without their knowledge or consent.
In addition to IUDs, the program included contraceptive injections and other interventions, largely disguised as other procedures. This crime against humanity continued through 1991, when Greenland finally wrested its health system from its colonial overlord’s control.
The new human-rights investigation heard from 354 women describing 488 separate incidents. Of 410 reported IUD insertions, 300 women said they had never consented. In another 63 cases, investigators had no consent information at all.
Nearly one-third of the women were 14 or younger at the time; more than half were minors. Some were just 12 years old.
These were not adults making informed medical decisions. They were schoolgirls summoned to clinics, fitted with devices they did not understand, and sent home in pain. Some bled. Some developed infections. Some suffered lasting physical and psychological damage. Some, like Petersen, lost the ability to bear children even after discovering what had been done.
One team of investigators found reasonable grounds to believe that Denmark’s conduct in imposing measures to prevent births met the definition contained in the Genocide Convention. The investigators stopped short of finding genocide because they could not establish specific intent for that exact crime. That distinction will matter in court, reducing possible claims against the Danish government. It does not rescue Denmark’s reputation.
Nor does Copenhagen seriously dispute the charge. In September 2025, Prime Minister Mette Frederiksen acknowledged that thousands of women and girls had received IUDs, many against their will. She admitted that Greenlanders had been subjected to systematic discrimination precisely because they were Greenlandic.
To her credit, she apologized on behalf of Denmark. But it’s far too little, far too late.
Denmark treated Greenlandic women not as patients or citizens, but as a herd to be culled. Petersen put it more plainly: “They took our resources. They took our bodies. And then they told us to thank them.”
Then Denmark Took The Children They Had
A state that believed it could decide which, if any, Greenlandic women would be allowed to become mothers also presumed to decide which Greenlandic mothers were allowed to keep their children.
The United Nations has reported that Inuit children from Greenland living in Denmark were seven times more likely than Danish-born children to be taken and placed outside the home. Danish authorities relied on standardized psychological and “parenting competency” tests developed for Danish parents, in the Danish language and within Danish cultural assumptions.
They then used those tests to judge Greenlanders whose language, emotional expression, extended-family structure and child-rearing traditions were profoundly different.
The result was both predictable and monstrous. Danish institutions defined Danish behavior as normal, interpreted Greenlandic standards as inferior, and took Greenlandic children from their families on that basis.
Mothers described being assessed without adequate interpreters. One says she was forbidden to speak Greenlandic to her twin sons during visits. The boys eventually lost their mother tongue as well as their mother.
It was a systematic effort to erase the Greenlandic population and culture for the benefit of an imperial master.
Denmark only abandoned the standardized tests for Greenlandic families in 2025, replacing them with assessments by a specialized unit with Greenlandic language and cultural competence.
Yet mere months later, authorities removed Ivana Nikoline Brønlund’s newborn daughter approximately one hour after birth without applying the new procedure. Denmark’s Social Minister called the failure a serious error. Yet only after widespread political outrage was the decision reversed, allowing mother and child to reunite.
This is not ancient history: this is yesterday.
Greenland Wants Independence. But It Can’t Afford It.
Greenlanders have been voting their way toward independence for nearly half a century. Understandably, more than 70 percent approved Home Rule in 1979. In 2008, 75.5 percent approved Self-Government — an arrangement that formally recognized Greenlanders as a people with the right to self-determination, and that established a legal path to full sovereignty.
By January 2025, a Verian poll found that 84 percent wanted independence, and 56 percent would vote for it even if the referendum were held immediately.
The hesitation is not the destination: it’s the price. Independence is expensive, and as badly as Denmark underfunds its colony, its colony doesn’t have the money to replace what it does spend. Independence means more than raising a flag.
A sovereign country must finance its government, build infrastructure, create and maintain embassies around the world, provide social services and defend an enormous, resource-rich island sitting astride the Arctic and Atlantic approaches to North America.
Greenland cannot do all of that alone. Denmark adds very little, but it’s far more than Greenland can do for itself.
So Greenland is not debating whether it should be free. It’s debating how to survive afterward.
The answer, of course, is a Compact of Free Association with the United States. It’s a model the U.S. has used successfully for and with Pacific nations such as Palau, Micronesia and the Marshall Islands. Each is a sovereign, independent nation. But each also receives many of the benefits of being part of the United States, defense included.
Citizens of these three COFA nations may live and work indefinitely in the United States, as nonimmigrants without a visa. They can qualify for American benefits ranging from Pell Grants and work-study programs to Medicare and Social Security. They may join the U.S. military. If disaster strikes, FEMA shows up. America invests and creates jobs in their countries.
A COFA provides all the benefits of inclusion in the world’s most powerful economy but without giving up independence. It also provides all the benefits of independence without the costs.
It’s genuinely “have your cake and eat it too.”
For Greenlanders, a COFA means independence without strategic helplessness. American capital and technology, far greater than anything Denmark can provide, would flow in to build the ports, airports, power systems, telecommunications, housing and processing facilities, all employing Greenland’s citizens and providing futures for its children.
Greenland’s vast deposits of rare earth elements, hydrocarbons and other minerals could create high-paying jobs and a booming local economy rather than leave the island dependent on a stingy annual subsidy. Its fisheries could be processed at home rather than shipped away so others capture the value. Its tourism — currently a fraction of Alaska’s — could make the country a fortune: just ask Iceland.
Greenland could even become the Arctic answer to Bermuda or the Caymans: a sovereign financial center inside the American economic system, with easy access to both New York and London. The nation could attract offshore wealth beyond anything it has ever imagined.
Perhaps even more important, Greenland's young people would no longer have to choose between dependence and emigration. Greenlanders could build a rich, modern country all their own.
So What Would America Get Out of This?
For America, the advantages are every bit as substantial. A free and friendly Greenland would secure the northern approaches to the United States, anchor the Greenland-Iceland-UK Gap, protect emerging Arctic sea lanes and enhance early warning, missile defense, space surveillance and North Atlantic monitoring.
A COFA would give American companies access to critical minerals outside China’s control while denying Beijing or Moscow the chance to turn Greenlandic poverty into a strategic foothold threatening the approaches to North America.
America does not need to actually own Greenland. It needs Greenland to be free, prosperous and able to choose a durable partnership with America. That is partnership, not colonialism: indeed, it would replace the current colonialism with a true partnership. Greenlanders would own Greenland, while both nations become richer and safer.
This solves Greenlanders’ long-term problem even more than America’s. But it also solves a problem that almost no one in today’s ahistorical political class grasps. Despite its shameful behavior, Denmark is America’s friend today. But no one can guarantee the future.
The peaceful North Atlantic order is a constructed alignment, not a law of nature. The structure of power and interests may change. The map does not. And leaders should not base the security of those entrusted to them on utopian dreams or “permanent” alliances and institutions.
The example of Diego Garcia is a stark reminder. America chose Diego Garcia, its most vital Indian Ocean base, in part because of Britain’s “unquestioned sovereignty” over the island. Britain remains one of America’s closest allies.
Yet Britain's socialist government has signed a treaty to transfer sovereignty over the Chagos Archipelago to Chinese-aligned Mauritius. It assures the United States and its own people that all may rely on “ironclad” treaty protections for continued American and British access to the base. Yet that’s obviously nonsense.
Britain did not have to become America’s enemy to alter the strategic arrangement on which our security relies. Governments change. Treaties change. Interests change. Alliances change. Geography does not.
Greenland will remain astride North America’s Arctic and Atlantic approaches no matter who governs in Copenhagen or Nuuk five, 50, or 100 years from now. Nations do not have the luxury of assuming every future government will play nice forever. American statesmen must plan for the geography their grandchildren will inherit, not merely for the diplomatic comfort of the present.
Denmark’s abuses answer the moral question. Greenlandic independence answers the colonial question. Free association with America answers the economic and strategic questions. It also enables what last week’s reports make clearly necessary, but which would be impossible without American help.
Denmark’s rule must end. Greenlanders must be free. But for that to happen, America alone can offer them a future in which independence means sovereignty, prosperity and security — not poverty, vulnerability or a new Chinese master.
The Senate math nobody on cable will say plainly:
Democrats need to NET FOUR seats. But 20 of 22 GOP seats up are in states Trump won by DOUBLE DIGITS. Their only real targets: Maine and North Carolina. Meanwhile THEY defend Georgia, Michigan, and New Hampshire.
They'd need to run the table perfectly and flip red states in a war economy. The map doesn't just favor us. It boxes them in.
Venezuela: The Biggest Oil Deal Ever — and Canada’s Strategic Folly
Cary Dunst and I discuss how Trump’s Venezuelan oil deal gives Washington new leverage over Ottawa, offers a free Iran a way to restore its energy wealth, and gives Venezuelans their futures back.
AI Just Created New Viruses. We Aren’t Ready.
Artificial intelligence has crossed the threshold from analyzing biology to creating entirely new viruses. An AI-created pandemic may be a weapon — or a catastrophic global accident.
You know I use and strongly support artificial intelligence. It is a boon to mankind and has increased my own productivity fivefold at least. It is an enormous blessing.
So is nuclear power, which is also the best way to power AI. What is not so great is a nuclear bomb, especially if you’re beside it when it goes off.
So understand that distinction when I sound this alarm. AI just crossed a line that can’t readily be uncrossed. It can now design complete, viable viruses that have never existed in nature, and for which no one has immunity.
Researchers at Stanford University and the Arc Institute trained an AI model called Evo on vast libraries of genetic material. They then taught it the genomes of Phi X-174 — a tiny virus that infects E. coli — and roughly 15,000 of its relatives. The model generated 700,000 candidate viral genomes. The scientists selected 285, manufactured their DNA, and inserted it into bacteria.
Sixteen worked. They reproduced. They burst out of their hosts. They infected other bacteria. Some multiplied faster than the natural virus on which the experiment was based.
To their credit, the researchers deliberately excluded viruses that infect humans and other animals, and even those that infect plants and fungi. The sixteen successful creations are bacteriophages. They are not going to turn into airborne smallpox and wipe out Chicago, or decimate humanity and create the Planet of the Apes.
But all that misses the point. Because while they won’t, something like them might.
Until now, the great promise of biological AI was that it could read the language of life: find patterns, predict protein structures, identify dangerous mutations, discover drugs, and help doctors repair what disease had broken. And indeed, the promise of that research is vast, and can and will revolutionize healthcare and longevity for everyone.
But now that same AI has gone beyond such benign research and created something entirely new — not merely a protein or a DNA fragment, but complete, viable viruses.
The experiment was safe. The capability is not.
AI’s peculiar strength is cheap failure at immense scale. Evo generated 700,000 possibilities; the researchers could rank them, manufacture the most promising, learn from the results, and try again. A pandemic pathogen does not require every design to work. It just needs one. And every failed round teaches the next model how to improve.
Bill Gates has now sounded the alarm. Last week, he told CNN’s Anderson Cooper and The Atlantic’s Hanna Rosin that today’s AI models are already “crossing the danger thresholds” in bioterrorism, that systems capable of helping create dangerous molecules are not adequately monitored, and that the next generation of models will be dramatically better.
He’s right. He’s also late.
The First One May Be an Accident
The Stanford team did everything responsibly. But Stanford isn’t the whole world. Its safety practices don’t bind a laboratory in China (obviously), Iran, Russia, or some deranged leftist depopulation fanatic at any of thousands of universities.
They don’t even bind the next American laboratory that persuades itself that the desired medical payoff justifies just one more experiment.
We’ve all seen that movie. Now it’s happening.
Nor is it unprecedented. Obvious to all at the time except the willingly self-deceived, the evidence is now overwhelming that COVID-19 was lab-created.
The Wuhan Institute of Virology was conducting risky work made possible in part by American taxpayer dollars routed through EcoHealth Alliance. Both the FBI and a congressional investigation found that the pandemic almost certainly originated there.
Finally, this summer, the Director of National Intelligence released additional evidence and stated directly that U.S.-funded gain-of-function research at Wuhan was the source of the accidental leak. Gee thanks, Tony Fauci!
We don’t need the missing laboratory notebook to understand the lesson. We don’t need to believe the scientists wanted to kill anyone. They probably didn’t. If they had, the pandemic probably would have been much worse.
But all of that is kind of irrelevant, isn’t it? Millions of people died either way. Economies shut down. Children lost years of education. Governments asserted powers they’d never possessed in peacetime. Supply chains shattered. Public trust collapsed. And all of it followed from research supposedly undertaken for defensible scientific reasons, by credentialed people you paid for at respected institutions.
That’s all without AI.
The first AI-created pandemic may come from a terrorist or a foreign military, but it doesn’t have to. All it will need is a grant, a safety committee, an ambitious scientist, and a single irreversible catastrophic mistake.
Every Military Will Want It
The problem is, terrorists, dictators, and militaries do exist.
Militaries exist to kill people and break things, to destroy an enemy’s capacity and will to resist. Sometimes that mission saves civilization. Sometimes it serves barbarism. Both inhabit the same world. And no competent general staff can ignore a weapon its enemies might acquire. Research isn’t optional: it’s required.
That is the remorseless logic of any arms race. If the United States believes China can design a pathogen, we don’t have the luxury of assuming the CCP’s benevolence, honesty, or competence. We have no choice but to develop the capability to match or defeat it. If China believes the U.S. is pursuing exactly that, it must do the same. Every serious power on Earth faces the same merciless calculation.
And there’s no question: that logic has averted immense harm. Despite the never-higher tensions of the Cold War, neither side started a nuclear war because, ultimately, they knew they couldn’t finish it: it would finish them.
A similar logic kept chemical weapons off European battlefields during World War II. The major powers had them, and prepared for them, but unlike in the prior war, never once used them. That’s pretty remarkable in the most destructive conflict in history, with combatants as evil as Hitler and Stalin. Yet chemical weapons weren’t used — deterrence worked — precisely because everyone had them.
As Heinlein said, an armed society is a polite society.
But deterrence requires the weapon to exist. It requires rival nations to build it, test it, store it, and entrust it to fallible human beings. It doesn’t deter accidents. It doesn’t deter a regime already facing destruction. It certainly doesn’t deter a terrorist who expects and longs to die. And unlike a missile, a virus is tiny, reproduces after launch, crosses borders without a passport, and may conceal both its origin and its attacker.
I can attest: AI is progressing at a staggering rate. If you use it every day like I do, you know exactly what I mean. It does new things, seemingly every day, that simply take my breath away. And the rate of advance is accelerating.
AI doesn’t eliminate the need for laboratories, specialized equipment, or skilled biologists. Rather, it compresses the search. Work that once demanded years of intuition, trial, error, and accumulated institutional knowledge can be proposed, compared, and refined in minutes or seconds.
That lowers the cost of competence. It allows small teams to attempt what once required vast programs with funding to match. I know, from personal experience.
Tomorrow’s systems will be even better (maybe literally tomorrow: major new upgrades now ship within weeks of each other). Those models will learn more of the grammar of biology, model more complicated organisms, and search possibility spaces no human or team of humans could examine in a thousand lifetimes.
Military planners will inevitably ask them to optimize transmissibility, lethality, incubation time, resistance to treatment, and host range. As the models learn more about human receptors, immune responses, and population genetics, someone will also ask whether a disease can be aimed at people who share particular inherited traits.
Imagine a bioweapon that only kills military-aged men. Or only kills pregnant women. Or only kills the specific race predominant in some enemy’s country. This is an apocalyptic movie. And it’s becoming possible.
Ten years ago, a machine that could carry on this conversation, write software, diagnose diseases, or design a viable virus would have been dismissed as science fiction, by pretty much everyone except Ray Kurzweil and Andrej Karpathy. The models aren’t going to stop improving, for better or for worse.
Can this nightmare be prevented? Possibly. But in the Cold War, deterrence worked in part because there were only two rival superpowers, and the cost of a nuclear weapons program was astronomical. AI puts incredible power in the hands of far more people at a far lower price. And therein lies the danger.
We’ve Seen This Play Before
After 1492, Europeans carried diseases into the Americas that tore through peoples with no previous exposure to many of them. Roughly 90 percent of the population of the hemisphere died. Entire societies were broken before conquerors could have defeated them by force of arms.
The Europeans did not invent those pathogens, or have any idea of the harm they would cause. The natives just lacked immunity, and the diseases did their work.
AI opens the possibility that human beings could manufacture such an immunological imbalance on purpose — creating a pathogen no population has encountered in precisely that form, then allowing evolution to continue the experiment after release. Human immunity is enormously complex; novelty alone does not guarantee catastrophe.
But no sane civilization would volunteer eight billion people as the test population (well, unless they were selling a COVID vaccine).
Nor can America solve this by passing an American law. A domestic rulebook can’t stop a lab overseas, hostile or merely sloppy. The Biological Weapons Convention prohibits these weapons, but it lacks the kind of standing inspection and verification machinery created for nuclear or chemical weapons.
That deficiency was a problem before AI. Now it’s downright insane.
The nations controlling frontier AI, advanced computing, DNA synthesis, laboratory equipment, and the global financial system must build an enforceable counterproliferation regime while they still can. It will never be perfect. But imperfection isn’t an argument for suicide.
Scientific freedom is precious, but it’s not a suicide pact. We license nuclear reactors, test medicines, inspect passenger aircraft, and imprison people who mishandle explosives. We do so because freedom doesn’t include the right to impose catastrophic risks on everyone else. In the beginning we didn’t know what we were doing, but we figured it out. We can again. We have to try.
But if it isn’t universal, it’s not going to work.
The Tools Are Good. But Man Is Fallen, Fallible, and Finite
This is definitely not an argument against artificial intelligence, biotechnology, or scientific progress. AI-designed phages may someday kill antibiotic-resistant bacteria that medicine cannot. Engineered viruses already deliver lifesaving genes into human cells. The same tools may cure inherited diseases, accelerate vaccines, and extend healthy human life. They also do a million other things having nothing to do with biology.
The technology isn’t evil. It’s powerful. And like every other technology, it’s in the hands of people who may be either good or evil, careful or foolish. And none of them know everything, or can see the future.
Scripture has warned us for millennia — and science fiction for generations — that human beings would remain stubbornly human even while acquiring godlike tools. The tools are neutral, like everything else in this world. Only the people are moral actors, using them for good or ill.
Stanford and the Arc Institute created sixteen viruses. They proved the design loop works. The models will improve. The costs will fall. The knowledge will spread. The military logic will take hold.
Bill Gates is sounding the alarm now. Every government better listen. We have crossed from reading the code of life to writing it. If civilization does not learn to govern that power before the next accident — or the next war — biology will deliver the verdict for us.
(I write here and on my site -- link in bio -- every day. Follow me here and subscribe over there.)
AI Just Created New Viruses. We Aren’t Ready.
Artificial intelligence has crossed the threshold from analyzing biology to creating entirely new viruses. An AI-created pandemic may be a weapon — or a catastrophic global accident.
You know I use and strongly support artificial intelligence. It is a boon to mankind and has increased my own productivity fivefold at least. It is an enormous blessing.
So is nuclear power, which is also the best way to power AI. What is not so great is a nuclear bomb, especially if you’re beside it when it goes off.
So understand that distinction when I sound this alarm. AI just crossed a line that can’t readily be uncrossed. It can now design complete, viable viruses that have never existed in nature, and for which no one has immunity.
Researchers at Stanford University and the Arc Institute trained an AI model called Evo on vast libraries of genetic material. They then taught it the genomes of Phi X-174 — a tiny virus that infects E. coli — and roughly 15,000 of its relatives. The model generated 700,000 candidate viral genomes. The scientists selected 285, manufactured their DNA, and inserted it into bacteria.
Sixteen worked. They reproduced. They burst out of their hosts. They infected other bacteria. Some multiplied faster than the natural virus on which the experiment was based.
To their credit, the researchers deliberately excluded viruses that infect humans and other animals, and even those that infect plants and fungi. The sixteen successful creations are bacteriophages. They are not going to turn into airborne smallpox and wipe out Chicago, or decimate humanity and create the Planet of the Apes.
But all that misses the point. Because while they won’t, something like them might.
Until now, the great promise of biological AI was that it could read the language of life: find patterns, predict protein structures, identify dangerous mutations, discover drugs, and help doctors repair what disease had broken. And indeed, the promise of that research is vast, and can and will revolutionize healthcare and longevity for everyone.
But now that same AI has gone beyond such benign research and created something entirely new — not merely a protein or a DNA fragment, but complete, viable viruses.
The experiment was safe. The capability is not.
AI’s peculiar strength is cheap failure at immense scale. Evo generated 700,000 possibilities; the researchers could rank them, manufacture the most promising, learn from the results, and try again. A pandemic pathogen does not require every design to work. It just needs one. And every failed round teaches the next model how to improve.
Bill Gates has now sounded the alarm. Last week, he told CNN’s Anderson Cooper and The Atlantic’s Hanna Rosin that today’s AI models are already “crossing the danger thresholds” in bioterrorism, that systems capable of helping create dangerous molecules are not adequately monitored, and that the next generation of models will be dramatically better.
He’s right. He’s also late.
The First One May Be an Accident
The Stanford team did everything responsibly. But Stanford isn’t the whole world. Its safety practices don’t bind a laboratory in China (obviously), Iran, Russia, or some deranged leftist depopulation fanatic at any of thousands of universities.
They don’t even bind the next American laboratory that persuades itself that the desired medical payoff justifies just one more experiment.
We’ve all seen that movie. Now it’s happening.
Nor is it unprecedented. Obvious to all at the time except the willingly self-deceived, the evidence is now overwhelming that COVID-19 was lab-created.
The Wuhan Institute of Virology was conducting risky work made possible in part by American taxpayer dollars routed through EcoHealth Alliance. Both the FBI and a congressional investigation found that the pandemic almost certainly originated there.
Finally, this summer, the Director of National Intelligence released additional evidence and stated directly that U.S.-funded gain-of-function research at Wuhan was the source of the accidental leak. Gee thanks, Tony Fauci!
We don’t need the missing laboratory notebook to understand the lesson. We don’t need to believe the scientists wanted to kill anyone. They probably didn’t. If they had, the pandemic probably would have been much worse.
But all of that is kind of irrelevant, isn’t it? Millions of people died either way. Economies shut down. Children lost years of education. Governments asserted powers they’d never possessed in peacetime. Supply chains shattered. Public trust collapsed. And all of it followed from research supposedly undertaken for defensible scientific reasons, by credentialed people you paid for at respected institutions.
That’s all without AI.
The first AI-created pandemic may come from a terrorist or a foreign military, but it doesn’t have to. All it will need is a grant, a safety committee, an ambitious scientist, and a single irreversible catastrophic mistake.
Every Military Will Want It
The problem is, terrorists, dictators, and militaries do exist.
Militaries exist to kill people and break things, to destroy an enemy’s capacity and will to resist. Sometimes that mission saves civilization. Sometimes it serves barbarism. Both inhabit the same world. And no competent general staff can ignore a weapon its enemies might acquire. Research isn’t optional: it’s required.
That is the remorseless logic of any arms race. If the United States believes China can design a pathogen, we don’t have the luxury of assuming the CCP’s benevolence, honesty, or competence. We have no choice but to develop the capability to match or defeat it. If China believes the U.S. is pursuing exactly that, it must do the same. Every serious power on Earth faces the same merciless calculation.
And there’s no question: that logic has averted immense harm. Despite the never-higher tensions of the Cold War, neither side started a nuclear war because, ultimately, they knew they couldn’t finish it: it would finish them.
A similar logic kept chemical weapons off European battlefields during World War II. The major powers had them, and prepared for them, but unlike in the prior war, never once used them. That’s pretty remarkable in the most destructive conflict in history, with combatants as evil as Hitler and Stalin. Yet chemical weapons weren’t used — deterrence worked — precisely because everyone had them.
As Heinlein said, an armed society is a polite society.
But deterrence requires the weapon to exist. It requires rival nations to build it, test it, store it, and entrust it to fallible human beings. It doesn’t deter accidents. It doesn’t deter a regime already facing destruction. It certainly doesn’t deter a terrorist who expects and longs to die. And unlike a missile, a virus is tiny, reproduces after launch, crosses borders without a passport, and may conceal both its origin and its attacker.
I can attest: AI is progressing at a staggering rate. If you use it every day like I do, you know exactly what I mean. It does new things, seemingly every day, that simply take my breath away. And the rate of advance is accelerating.
AI doesn’t eliminate the need for laboratories, specialized equipment, or skilled biologists. Rather, it compresses the search. Work that once demanded years of intuition, trial, error, and accumulated institutional knowledge can be proposed, compared, and refined in minutes or seconds.
That lowers the cost of competence. It allows small teams to attempt what once required vast programs with funding to match. I know, from personal experience.
Tomorrow’s systems will be even better (maybe literally tomorrow: major new upgrades now ship within weeks of each other). Those models will learn more of the grammar of biology, model more complicated organisms, and search possibility spaces no human or team of humans could examine in a thousand lifetimes.
Military planners will inevitably ask them to optimize transmissibility, lethality, incubation time, resistance to treatment, and host range. As the models learn more about human receptors, immune responses, and population genetics, someone will also ask whether a disease can be aimed at people who share particular inherited traits.
Imagine a bioweapon that only kills military-aged men. Or only kills pregnant women. Or only kills the specific race predominant in some enemy’s country. This is an apocalyptic movie. And it’s becoming possible.
Ten years ago, a machine that could carry on this conversation, write software, diagnose diseases, or design a viable virus would have been dismissed as science fiction, by pretty much everyone except Ray Kurzweil and Andrej Karpathy. The models aren’t going to stop improving, for better or for worse.
Can this nightmare be prevented? Possibly. But in the Cold War, deterrence worked in part because there were only two rival superpowers, and the cost of a nuclear weapons program was astronomical. AI puts incredible power in the hands of far more people at a far lower price. And therein lies the danger.
We’ve Seen This Play Before
After 1492, Europeans carried diseases into the Americas that tore through peoples with no previous exposure to many of them. Roughly 90 percent of the population of the hemisphere died. Entire societies were broken before conquerors could have defeated them by force of arms.
The Europeans did not invent those pathogens, or have any idea of the harm they would cause. The natives just lacked immunity, and the diseases did their work.
AI opens the possibility that human beings could manufacture such an immunological imbalance on purpose — creating a pathogen no population has encountered in precisely that form, then allowing evolution to continue the experiment after release. Human immunity is enormously complex; novelty alone does not guarantee catastrophe.
But no sane civilization would volunteer eight billion people as the test population (well, unless they were selling a COVID vaccine).
Nor can America solve this by passing an American law. A domestic rulebook can’t stop a lab overseas, hostile or merely sloppy. The Biological Weapons Convention prohibits these weapons, but it lacks the kind of standing inspection and verification machinery created for nuclear or chemical weapons.
That deficiency was a problem before AI. Now it’s downright insane.
The nations controlling frontier AI, advanced computing, DNA synthesis, laboratory equipment, and the global financial system must build an enforceable counterproliferation regime while they still can. It will never be perfect. But imperfection isn’t an argument for suicide.
Scientific freedom is precious, but it’s not a suicide pact. We license nuclear reactors, test medicines, inspect passenger aircraft, and imprison people who mishandle explosives. We do so because freedom doesn’t include the right to impose catastrophic risks on everyone else. In the beginning we didn’t know what we were doing, but we figured it out. We can again. We have to try.
But if it isn’t universal, it’s not going to work.
The Tools Are Good. But Man Is Fallen, Fallible, and Finite
This is definitely not an argument against artificial intelligence, biotechnology, or scientific progress. AI-designed phages may someday kill antibiotic-resistant bacteria that medicine cannot. Engineered viruses already deliver lifesaving genes into human cells. The same tools may cure inherited diseases, accelerate vaccines, and extend healthy human life. They also do a million other things having nothing to do with biology.
The technology isn’t evil. It’s powerful. And like every other technology, it’s in the hands of people who may be either good or evil, careful or foolish. And none of them know everything, or can see the future.
Scripture has warned us for millennia — and science fiction for generations — that human beings would remain stubbornly human even while acquiring godlike tools. The tools are neutral, like everything else in this world. Only the people are moral actors, using them for good or ill.
Stanford and the Arc Institute created sixteen viruses. They proved the design loop works. The models will improve. The costs will fall. The knowledge will spread. The military logic will take hold.
Bill Gates is sounding the alarm now. Every government better listen. We have crossed from reading the code of life to writing it. If civilization does not learn to govern that power before the next accident — or the next war — biology will deliver the verdict for us.
(I write here and on my site -- link in bio -- every day. Follow me here and subscribe over there.)
🇨🇳🇻🇪🇺🇸 Chinese Foreign Ministry Spox.:
"China-Venezuela cooperation is protected by international law and the laws of both countries.
The cooperation does not target any third party and should not be subjected to interference by any third party.
China's lawful rights and interests in Venezuela must be protected."
Before the U.S.-Venezuela oil agreement, Chinese and Russian state companies HEAVILY managed and relied on Venezuelan oil fields...
Now, the U.S. essentially "runs" Venezuela's oil market, undercutting China's energy security (again).
Writer: Michael
At least 20 generals, admirals and civilian defense leaders have been fired or have departed their jobs since Pete Hegseth became defense secretary at the beginning of the second Trump administration. https://t.co/wwCnrwF10v
Venezuela: The Biggest Oil Deal Ever — and Canada’s Strategic Folly
Cary Dunst and I discuss how Trump’s Venezuelan oil deal gives Washington new leverage over Ottawa, offers a free Iran a way to restore its energy wealth, and gives Venezuelans their futures back.
@MarioNawfal Wouldn't Carney have to have some strength for "strength and directness" to work? But leftists truly believe they can conjure their own realities with wishes and unicorn farts.