The most dangerous technology of our age is not the one that replaces human labor, but the one that quietly replaces human solidarity — and nowhere is this substitution more consequential than in the life sciences, where the very biology of citizenship is at stake.
The Western political imagination has long rested on a fiction so foundational that we rarely name it: that the body politic is composed of bodies that matter equally. This fiction was never fully realized — racial hierarchy, class exclusion, and gendered citizenship have always scarred the promise of liberal democracy. Yet the aspiration itself, however imperfectly pursued, provided a common horizon. Citizens could disagree about the distribution of goods and rights precisely because they shared a substrate of mutual intelligibility: a common world of facts, a common biology, a common mortality.
That substrate is now being privatized. And the privatization is occurring most consequentially not in the noisy theater of electoral politics or social media, but in the quieter, more intimate domain of the life sciences — the domain that determines who lives, who suffers, and who is rendered statistically legible.
The New Sovereigns
Jack Balkin of Yale Law School has argued that we are witnessing the emergence of an „algorithmic state“ in which private technology companies increasingly provide the infrastructure through which government itself operates . Palantir governs foreign intelligence; Amazon governs cloud infrastructure; OpenAI and Anthropic govern the computational substrate of administrative decision-making . This is not a conspiracy but a structural condition: governments lack the technological capacity to govern without the platforms, and the platforms lack the constitutional obligations that bind governments.
Balkin calls the result „mutual co-optation“ — governments seek technological capacity from private firms, while the firms seek contracts, favorable regulation, and protection . The public-private distinction, on which so many constitutional protections depend, dissolves into a haze of contractual relationships and algorithmic opacity.
Balkin’s diagnosis of „constitutional rot“ identifies four horsemen: widening wealth inequality, political polarization, loss of trust, and policy disasters that leave citizens viewing government as incompetent or indifferent . He describes the present as a „second Gilded Age,“ marked by rapid technological disruption, extraordinary concentrations of wealth, bitter political division, and distrust of institutions . But his analysis, focused as it is on the constitutional order, understates the biological dimension of this transformation. The algorithmic state does not merely govern speech, commerce, and civic life. It increasingly governs the body.
The Life Sciences as Critical Infrastructure
Healthcare AI is often discussed as a question of efficiency — faster diagnoses, optimized treatment plans, streamlined administrative work. This framing obscures the deeper political stakes. The life sciences are not merely another sector to be disrupted. They are the infrastructure through which a society reproduces itself biologically. They determine who is born, who is cured, who is managed, and who is left to die.
When this infrastructure is privatized, the consequences are not merely economic. They are biopolitical in the most literal sense: they reshape the boundaries of the human community.
The World Economic Forum has documented that most AI benefits are confined to high-income economies, with AI capital, data, and compute increasingly controlled by a small number of firms and countries . Less than 1% of AI investment is directed toward social impact . Most AI systems are trained on just 100 of the world’s more than 7,000 languages . This divergence, the WEF argues, „signals not a failure of innovation, but of alignment“ .
The United Nations Secretary-General has delivered an even starker warning. In his remarks to the first Global Dialogue on AI Governance, he identified three warnings from the science. The second is about power: „The computing power, the data and the talent behind the most advanced systems are concentrated in a handful of companies, and in a handful of countries. Most nations – including many developing countries – have had no say in decisions that will shape their futures. The longer we wait, the harder that concentration sets. When power imbalances are hard-wired into technology, inequality becomes part of the code“ .
The UN’s Independent International Scientific Panel on AI, comprising 40 international experts, has reinforced this finding. Brazilian panel member Teresa Ludermir warned that AI development is concentrated in few countries and companies, resulting in „a concentration of computational power, data and talent“ that „can generate technological dependence and limit the capacity of developing countries“ . The panel’s report highlights AI’s potential benefits in medicine — predicting structures of over 200 million proteins, accelerating drug discovery and vaccine development — while simultaneously warning that these benefits risk deepening inequality .
The Biopolitics of Algorithmic Fragmentation
The fragmentation of the Western body politic through privately held AI will not announce itself as a political rupture. It will manifest as a series of biological gradients.
First gradient: Diagnostic legibility. If AI diagnostic systems are developed and validated primarily on data from affluent, predominantly white populations, then the bodies of the poor, the racialized, and the geographically marginalized will become less legible to the medical system.
This is not a hypothetical. A 2026 MICCAI study on dermatological AI found that skin-cancer triage systems, when calibrated for sensitivity, systematically over-predict on darker skin: at a fixed operating point, a specificity deficit of 11.0 percentage points on darker-skin patients was measured, translating to an estimated 70 excess unnecessary referrals per 1,000 darker-skin patients. The cause is structural: lesion-background contrast is physically lower on darker skin, producing a 5.2× signal-to-noise ratio reduction from lighter to darker tones .
The underlying data problem is severe. A comprehensive review of dermatological datasets found that across all major repositories analyzed in a Lancet Digital Health systematic review, merely 1.3% of images include ethnicity data and only 2.1% report Fitzpatrick skin type classification. Most critically: only ten images globally from Fitzpatrick type V individuals and a single image from a type VI individual were found across all analyzed datasets . The Fitzpatrick17k dataset, one of the most widely used benchmarks, contains only 2.4% type 6 images — the darkest skin — and models trained on it show significantly degraded performance for that group .
Second gradient: Therapeutic access. The most advanced AI-driven therapeutic development is being built and controlled by private entities. In March 2026, researchers published in Cell the results of LUMI-lab, a „self-driving“ laboratory that integrates a foundation model pretrained on over 28 million molecular structures with robotic automation. Through ten iterative cycles, LUMI-lab synthesized and evaluated more than 1,700 lipid nanoparticles, autonomously discovering „brominated lipid tails“ as a previously unrecognized structural feature enhancing mRNA delivery . The best-performing lipid, LUMI-6, achieved 20.3% epithelial cell editing efficacy in the lung — surpassing previously reported LNP-mediated CRISPR-Cas9 delivery via inhalation .
This is a genuine breakthrough. But the research team has filed a patent for LUMI-lab, including the model and ionizable lipids . The question of who benefits from such breakthroughs is not answered by the technology itself. It is answered by the ownership structures, pricing models, and patent regimes that surround it.
Senator Bernie Sanders has articulated a principle that the life sciences make urgent. His proposed American AI Sovereign Wealth Fund Act would impose a one-time 50% tax on the stock of AI companies reaching $200 million in annual AI sales, creating a fund Sanders estimates at nearly $7 trillion. A seven-person independent commission would manage the fund and use its voting shares „to block decisions that hurt the American people and to push for policies that help them.“ A 5% annual dividend would provide direct payments of more than $1,000 to every American .
„AI is built on the basis of human knowledge, built on the work of millions of people,“ Sanders said. „The benefits cannot simply go to the handful of wealthy corporations. They will be shared by the American people“ .
Sanders acknowledged that Sam Altman has expressed openness to public equity in AI companies, but they remained far apart on the scale. „I think people like Sam Altman and Trump (who) may be sympathetic to this are saying: ‚Okay, look, we’re making zillions of dollars so we’re going to be nice guys and maybe we’ll buy off the public. We will give 5% of our profits back into the government.‘ That’s not what we’re talking about“ .
Third gradient: Data colonialism. The extraction of health data from populations that will never benefit from the resulting therapies is not a metaphor. It is a business model. Academic analysis has identified „data colonialism“ as the process by which personal user data is collected, marketed, and misused for generating money — a pattern that „reminds one of the colonial patterns in the history of extracting resources“ . As algorithms are predesigned by the West, the effect is to deepen existing economic dependencies and project global disparities .
The Inter-Parliamentary Union has called for democratic leadership on AI, warning that „the current trajectory of AI, and the levels of risk associated with it, do not represent a democratic choice“ . The IPU’s signatories — including parliamentary leaders from Tanzania, Malaysia, Australia, Ecuador, the United Kingdom, and Uruguay — argue that „the risks that society accepts – and the boundaries that AI must respect – must be decided democratically“ .
„We do not accept the concentration of power in the hands of a few actors,“ the IPU declaration states. „Controls on AI do not mean a choice of safety over innovation. Societies have repeatedly decided that technology must operate within boundaries to protect the public. Aircraft must meet safety standards. Pharmaceuticals must be tested before reaching patients. Those requirements did not stop progress; they made the products safer and more trusted“ .
The Political Consequences
What happens to a democracy when its citizens no longer share a common biological reality?
The question is not abstract. It is the logical endpoint of the trends documented above.
If diagnostic AI works differently for different populations — as the dermatological evidence demonstrates — then „evidence-based medicine“ becomes a stratified epistemology. What counts as a valid diagnosis, a proven treatment, a legitimate clinical trial will depend on which body is being examined. The universal claims of science will be replaced by a patchwork of data-driven particularisms.
The pulse oximetry case illustrates the stakes. A Johns Hopkins study of 7,126 COVID-19 patients found that occult hypoxemia — dangerously low blood oxygen undetected by standard pulse oximeters — occurred in 28.5% of Black patients compared with 17.2% of White patients. Compared with White patients, SpO2 overestimated SaO2 by an average of 1.2% among Black patients. Among patients who should have been recognized as eligible for treatment, Black patients had a 29% lower hazard of treatment eligibility recognition. A total of 451 patients never had their eligibility recognized — most of whom (54.8%) were Black . Black patients who were eventually recognized faced a median delay of one hour longer than White patients .
This is not a bug in a single device. It is the logical output of systems that measure what is already stratified.
If therapeutic AI is privately owned and priced for profit, then the benefits of biomedical innovation will accrue to those who can pay. The rhetoric of „AI for all“ will collide with the reality of AI for those who can afford it. The result is not merely inequality. It is a biological gradient of citizenship — a world in which the quality and duration of life correlate directly with position in the social hierarchy.
The Optum algorithm case, documented in Science and analyzed in The Lancet Digital Health, demonstrates how this gradient operates even when race is ostensibly excluded. The algorithm assigned risk scores based on total healthcare costs, assuming higher costs indicated greater need. But Black patients generated an average of $1,800 less in annual healthcare costs than White patients with the same number of chronic conditions. The result: the algorithm failed to identify less than half the number of Black patients at risk of complicated medical needs as White patients . As The Lancet Digital Health editorial put it: „There is no such thing as race in health-care algorithms“ — only racism, embedded in the choice of proxy variables .
Constitutional Rot and Biological Secession
Balkin’s „constitutional rot“ framework becomes more alarming when extended to the biological domain. The four horsemen — widening inequality, polarization, loss of trust, policy disasters — are not merely political phenomena. They are biological gradients in the making.
Widening inequality is not merely a matter of income. It is a matter of life expectancy, of disease burden, of the biological capacity to flourish. Political polarization is not merely a matter of opinion. It is a matter of whether citizens share a common biology of hope. Loss of trust is not merely a matter of confidence in institutions. It is a matter of whether the institutions that govern health can be believed when they speak about bodies that are not white, not male, not wealthy.
The research literature on trust in healthcare AI confirms this dynamic. A consensus study published in the Journal of Medical Internet Research found that „personal and collective experiences with discrimination or degradation—along lines of race, class, gender, or other personal characteristics—especially create reasons for suspicion if not outright distrust“ . The study noted that trust may be „hard to achieve given prevailing and justified distrust in health service providers such as Optum and private health systems and marginalized communities‘ distrust in health care systems due to historic or entrenched disparities in outcomes or experiences of care“ .
The IPU’s warning is blunt: „With social media, we have seen what happens when a technology becomes embedded across society before its wider consequences are understood. With AI, people see extraordinary change being driven by powerful private interests, with unknown consequences for energy, jobs, information and privacy“ .
The Question of Ownership
At root, the question is one of ownership. Who owns the biological commons? Who owns the collective intelligence from which AI-driven medicine is built? Who owns the algorithms that determine diagnostic legibility, therapeutic access, and the distribution of life itself?
The current trajectory — private ownership, proprietary models, extracted data, stratified benefits — points toward a future in which the Western body politic fragments along biological lines. The rich and the poor will not merely live in different neighborhoods. They will inhabit different biological realities, mediated by different AI systems, validated by different data, governed by different epistemologies of the body.
This is not a future that technology determines. It is a future that ownership structures determine. The choice is between a medicine that reinforces the biological solidarity of the species and a medicine that dissolves it into privatized gradients of life and death.
The life sciences are critical infrastructure. To treat them as merely another sector for private disruption is to mistake the stakes. What is at stake is not efficiency. It is whether the West can remain a body politic at all — or whether it will dissolve into a collection of privately governed biologies, each with its own algorithms, its own evidence, its own truth about what a body is and whose body matters.
