Measuring AI-Induced Disempowerment: A Framework and Proposed Metrics

2026-06-25 · Source: Paper Index on ACL Anthology · Field: Science & Research — Social Sciences & Behavioral Studies, Research Methodology & Innovation · Depth: Expert, quick

Summary

A new framework and proposed metrics address the urgent need to measure AI-induced disempowerment, which tracks whether AI integration erodes humans' ability to meaningfully shape outcomes. The research operationalizes disempowerment using Sen's model of agency and a three-layer model encompassing exposure, erosion, and lock-in, applied across economic, political, and cultural domains at individual, institutional, and civilizational scales. Current measurement efforts are shown to cluster almost entirely at exposure, neglecting erosion and lock-in. Six concrete metrics are proposed: centaur evaluations, disempowerment perception surveys, AI content saturation and cultural convergence monitoring, monitoring capital flow to and from human labor, human task frontier tracking, and institutional ethnography. The paper also identifies key actors for implementation and discusses limitations like construct validity and causal attribution.

Key takeaway

For Policy Makers and Research Scientists developing AI impact assessments, recognize that current metrics often overlook AI-induced erosion and lock-in of human agency. You should integrate the proposed three-layer model and six concrete metrics, such as centaur evaluations and cultural convergence monitoring, to comprehensively track disempowerment across economic, political, and cultural domains. This ensures a more robust understanding of AI's societal effects and informs more responsible AI governance.

Key insights

Measuring AI-induced disempowerment requires a multi-layered framework and specific metrics beyond mere exposure.

Principles

• Disempowerment can be operationalized via Sen's agency model.
• AI's impact spans exposure, erosion, and lock-in layers.
• Measurement must cover economic, political, and cultural domains.

Method

The paper proposes a research agenda to measure AI-induced disempowerment using a three-layer model (exposure, erosion, lock-in) and six concrete metrics, identifying actors best positioned to implement each.

In practice

• Implement centaur evaluations for human-AI collaboration.
• Conduct disempowerment perception surveys.
• Monitor AI content saturation and cultural convergence.

Topics

• AI Disempowerment
• Human Agency
• AI Impact Measurement
• Ethical AI
• Societal AI Effects
• AI Governance

Best for: AI Scientist, AI Ethicist, Policy Maker, Research Scientist

Related on AIssential

• AI is becoming a second brain at the expense of your first one — Over-reliance on AI can erode human judgment, leading to disempowerment and the adoption of AI's biases.
• Measuring What Matters in the Age of AI Agents — Measure AI agent impact by treating them as team extensions, focusing on hybrid human-AI team performance.
• 19 Anti-Populist Takes on AI — AI's societal integration is complex, challenging simplistic narratives about human-machine conflict and its impact on labor, art, education, and regulation.
• A Framework for Measuring Appropriate Reliance on Set-Valued AI Advice — This framework is the first to formally measure appropriate reliance on set-valued AI advice in human-AI collaboration.
• Weekly Review 27 February 2026 — AI's pervasive impact spans ethical dilemmas, job displacement, environmental concerns, and technical advancements.
• "This Wasn't Made for Me": Recentering User Experience and Emotional Impact in the Evaluation of ASR Bias — ASR bias creates significant emotional and cognitive burdens for users of non-dominant dialects, extending beyond mere accuracy issues.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Paper Index on ACL Anthology.