Sheila Jasanoff

In the AI Story

Jasanoff's intellectual biography is a study in disciplinary crossing. Mathematics gave her precision, linguistics gave her sensitivity to how language shapes thought, and law gave her the institutional literacy to understand how power operates through rules that present themselves as neutral. Her 1990 book The Fifth Branch: Science Advisers as Policymakers established her reputation by demonstrating that scientific advisers do not merely inform policy — they make it, through the framing choices they impose on problems and the evidentiary standards they apply to solutions. The book revealed that regulatory science is not applied science but a hybrid form of knowledge, co-produced by scientific methods and institutional contexts.

The concept of co-production emerged from this work and has become Jasanoff's most influential contribution. Scientific knowledge and social order are not independent — they are made together. When a new scientific claim is established, it does not merely add to the stock of knowledge. It reorganizes the social world: creating new categories of people, new institutions, new hierarchies of authority, new configurations of power. Simultaneously, the social order shapes the science — what gets studied, what gets funded, what questions are deemed worth asking. This framework dissolves the fiction that technology develops autonomously and governance reacts. Both are being produced in the same process, and governing technology means governing the social order it co-produces.

Her comparative work on biotechnology governance across the United States, United Kingdom, Germany, and the European Commission introduced civic epistemology — the culturally embedded ways different societies produce and validate knowledge claims. The American civic epistemology privileges adversarial contestation and empirical demonstration. The German privileges consensual deliberation and precautionary assessment. These are not merely different regulatory postures but different ways of knowing, and they produce different relationships between experts and citizens, different evidentiary standards, and different governance outcomes. Applied to AI, civic epistemology explains why global governance conversations consistently produce mutual incomprehension: participants are not merely disagreeing about policy; they are operating within incommensurable frameworks for what counts as evidence and what constitutes legitimate authority.

Her concept of technologies of humility — introduced in a 2003 essay that has become one of the most cited works in science and technology studies — provides the prescriptive counterpart to her analytical framework. Humility is not modesty but an institutional capacity: the recognition that governance under uncertainty requires practices designed to acknowledge what cannot be known, to incorporate diverse knowledge, and to detect emergent consequences that no prediction could anticipate. The four components — framing, vulnerability analysis, distributional inquiry, and learning mechanisms — together constitute an institutional posture capable of governing technologies whose most important consequences are uncertain in the technical sense: not merely unknown but unknowable in advance, emerging from interactions no model can capture.

Origin

Jasanoff was born in Calcutta in 1944 and emigrated to the United States as a child. Her academic formation bridged continents and disciplines in ways that shaped her intellectual method. At Harvard as an undergraduate she studied mathematics and developed the analytical rigor that characterizes her work. In Germany she studied linguistics and absorbed the European intellectual tradition's attention to how language shapes reality. At Harvard Law School she learned how institutions work — how rules are made, how power operates through them, how procedural choices determine substantive outcomes. This combination of formal training and cross-cultural experience prepared her to see what specialists could not: that the governance of science and technology is not a technical problem requiring better expertise but a democratic problem requiring institutional redesign.

Key Ideas

Co-production of knowledge and social order. Scientific knowledge and social arrangements are made simultaneously, each constituting the other. When AI enters a workplace, it does not merely change what workers produce — it changes what a worker is, what expertise means, and who possesses authority.

Civic epistemology determines governance outcomes. Different societies validate knowledge through different institutional practices — adversarial demonstration, consensual deliberation, strategic assessment — and these differences produce incompatible governance frameworks that cannot be resolved through technical argument alone.

Technologies of humility as governance practice. Governing powerful technologies under conditions of genuine uncertainty requires institutional practices that frame problems broadly, analyze vulnerability systematically, examine distributional consequences explicitly, and build learning mechanisms that detect emergent harms.

Expert knowledge is necessary but insufficient. The people who build technologies possess knowledge essential to governance, but their knowledge is partial. Affected communities possess experiential knowledge that governance frameworks systematically exclude, and that exclusion produces technically competent but democratically illegitimate decisions.

Legitimacy requires participation. Governance adequate to AI demands institutional redesign so that the people whose lives are reshaped by the technology participate in decisions about that reshaping — not as beneficiaries of others' choices but as co-authors of their own future.

Debates & Critiques

Jasanoff's framework has been criticized from multiple directions. Technology optimists argue that her emphasis on precaution and participation slows innovation and privileges talk over action. Some policy scholars contend that participatory governance is too slow and too expensive for technologies that develop as rapidly as AI. Critical theorists argue that her framework accepts the legitimacy of capitalist markets and state institutions rather than interrogating them. Defenders respond that Jasanoff's work provides the most sophisticated account of how democratic societies can govern powerful technologies without surrendering either technical competence or democratic legitimacy, and that the tension she identifies between speed and participation is real but navigable through institutional design.