Decomposing Autonomy: Explaining AI Technology Acceptance Through a Liberty-Based Framework
This is a conceptually ambitious CHI paper that earns its place by making autonomy more precise: positive liberty, negative liberty, and agency are separated and then tied to AI acceptance. The empirical study is not huge, but it is enough to support the framework’s basic plausibility and to surface a non-obvious asymmetry between the two liberty dimensions.
Axes Lens
Rare contribution shape, typical evidence profile. The point here is not a score. It is to show what kind of claim the paper makes, and whether the evidence pattern is unusual or baseline in this 268 -review set.
Contribution shape
- Knowledge form
- descriptive knowledge typical · 92/268
- Novelty type
- framework typical · 59/268
- Abstraction level
- field typical · 41/268
- Generalization target
- field argument typical · 55/268
- Validation mode
- mixed methods typical · 136/268
Evidence profile
- Evidence strength
- moderate typical · 105/268
- Claim alignment
- strong typical · 231/268
- Overclaim risk
- medium typical · 210/268
Review Summary
The paper’s strongest contribution is not a new interface or algorithm, but a reframing move that is genuinely useful for CHI: it decomposes autonomy into positive liberty, negative liberty, and agency, borrowing a distinction from political philosophy and turning it into an HCI-relevant explanatory framework. That is a meaningful departure from the common-sense treatment of autonomy as a single, vaguely positive good. The abstract and results indicate that the authors do more than speculate: they run an online vignette study with 194 participants and show that positive and negative liberty are correlated yet distinct, with different relationships to agency and willingness to use technology. The reported negative effect of negative liberty on willingness to use is especially interesting because it resists the naive expectation that more freedom from constraints should always improve acceptance. At the same time, the paper’s own limitations matter for how far the claims can travel. The vignette design did not fully realize the intended contrasts, the scenarios are productivity-oriented AI-mediated communication cases, and the constructs capture subjective perceptions rather than objective conditions. So the paper is best read as a field-level conceptual and empirical argument: it offers a sharper vocabulary and a plausible explanatory model, but not a universal design prescription. In CHI terms, that makes it valuable as a synthesis/framework paper with moderate empirical support and clear boundary conditions, rather than as a broadly generalizable behavioral law.
What Changed
Canon before
Autonomy in HCI and AI acceptance is often treated as a broad, somewhat vague construct, frequently overlapping with agency, control, and freedom. Prior work commonly emphasizes user control or freedom from interference without decomposing autonomy into philosophically distinct components.
Departure from common sense
The paper reports that higher negative liberty does not simply increase acceptance; instead, the measured effect on willingness to use is slightly negative. That is a useful corrective to the intuitive assumption that more freedom from constraints always makes AI more acceptable.
Actual novelty
The paper’s main contribution is a liberty-based decomposition of autonomy that imports positive and negative liberty from political philosophy and maps them onto AI acceptance. It treats autonomy as a structured construct rather than a single undifferentiated preference, and uses that decomposition to explain willingness to use technology and sense of agency.
Evidence
The paper argues for and empirically tests a conceptual framework that separates positive liberty, negative liberty, and agency. The evidence includes an online vignette study with 194 participants, factor analyses showing PL and NL as related but separable constructs, regression/SEM results showing PL predicts intention to use while NL predicts agency, and explicit limitations about vignette design, scenario scope, and subjective-perception measurement.
“ This paper disentangles autonomy by integrating the dualistic nature of positive and negative liberty from the perspective of political philosophy”
actual novelty · Share on · confidence 0.80
“Decomposing Autonomy: Explaining AI Technology Acceptance Through a Liberty-Based Framework | Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems”
departure from common sense · Results, 5.3 Predicting Intention to Use · confidence 0.78
“ First, the design of vignettes did not fully reflect our theoretical rationale on PL- and NL-dominated scenarios with limited level contrasts for one of them”
limitation · 6.5 Limitations & Future Work · confidence 0.95
“ Using an online vignette study with N=194 participants, we show that positive and negative liberty act as correlated but distinct dimensions of the autonomy foundation”
validation scope · Abstract / Introduction · confidence 0.92
Limits
Method limits
The authors note that the vignette design did not fully reflect the intended PL- and NL-dominated contrasts, that the study relies on hypothetical scenarios rather than lived experience, and that the manipulations produced limited level contrasts in some conditions.
Deployment limits
The empirical setting is limited to AI-mediated communication scenarios with a productivity goal, so direct transfer to other AI domains, tasks, or organizational settings should be cautious.
Boundary conditions
Findings are bounded by scenario framing, limited contrast in the vignette manipulations, and the distinction between perceived liberty and objective liberty conditions. The reported effects should be read as evidence about acceptance in hypothetical communication contexts rather than universal design rules.
Position in field
This work sits at the intersection of HCI, AI acceptance, and political-philosophy-informed conceptual modeling. Its value is less in a new system artifact than in reframing autonomy as a decomposable explanatory construct that can sharpen future design and evaluation discussions.