Anticipatory Governance

The Core Idea

Anticipatory governance means asking governance questions while a technology is still being shaped, tested, funded, or justified rather than waiting until it is already embedded in everyday life.

The point is not to predict the future perfectly. The point is to ask hard questions while meaningful choices are still open.

Why Timing Matters

Early governance is difficult because evidence is incomplete. People do not yet know exactly how a system will be used, who will adopt it, or which harms will become most visible.

Late governance is difficult for a different reason. By the time harms are obvious, the system may already be tied to budgets, contracts, staffing patterns, institutional routines, technical dependencies, and habits of use. What once looked optional can start to feel inevitable.

Anticipatory governance asks us to work inside that uncomfortable window: early enough to shape the system, but seriously enough to ask what should happen before the path hardens.

What Anticipatory Governance Asks

• What futures are being imagined or promised?
• Who defines the problem the technology is supposed to solve?
• Who benefits from fast deployment?
• Who might bear the risks or costs?
• What could become hard to undo later?
• What alternatives are being ignored or displaced?
• Who needs to be involved before deployment?
• What forms of oversight, contestability, repair, or refusal should exist from the start?

Use This Concept When…

Use anticipatory governance when:

• a technology is being piloted, funded, tested, or scaled
• a system is described as inevitable or too important to slow down
• harms are treated as something to address later
• a proposed system shifts surveillance, institutional power, decision-making, or accountability
• technical performance claims are being used to justify deployment before public questions have been answered

Practice Using the Framework

Practice using this framework: How to Use Anticipatory Governance

Where This Idea Comes From

Anticipatory governance is associated with David Guston and related work on upstream governance, foresight, reflection, and public engagement. The central claim is that emerging technologies should not be governed only after they are widely deployed and politically entrenched.

This work argues that governance should happen while a technology is still taking shape: when goals, constraints, safeguards, and public commitments can still be revised.

Erik Fisher’s idea of midstream modulation adds another important point. Governance does not begin only after a product is released. It can also happen during research and development, when researchers and engineers are still deciding what to optimize, what to measure, what to test, and what kinds of use they are building toward.

Related Concepts

• Accountability: Automation changes accountability
• Surveillance: Being watched changes what counts as normal behavior
• Anomaly detection: Anomaly Detection
• Procedural justice: Procedural Justice
• Sociotechnical systems: AI systems are socio-technical systems