Answers are grounded in this paper and cite their sources.
AI legal assistants may expand preliminary information access, but perceived usefulness and willingness to consider a service should not be equated with improved access to justice. This study examines their alignment using 385 adult, consenting, AI-aware responses from Bangladesh, collected through a mixed online and in-person questionnaire approach. The legally oriented sample includes 227 law students. Low internal consistency across four questionnaire blocks motivates item-level analysis rather than composite mediation. Methods combine full response distributions, paired endorsement contrasts, joint response patterns, rank correlations and adjusted binary and ordinal models. Basic access, time saving and convenience have identical means of 3.85, yet 44.9% of respondents differ in endorsement between time saving and basic access. Willingness to consider preliminary AI guidance is endorsed by 63.9%; 30.4% jointly endorse time saving, convenience and willingness. The conditional statement that transparent explanations increase willingness receives 87.5% endorsement, exceeding general willingness by 23.6 percentage points. No focal coefficient survives multiplicity correction in the primary adjusted binary model; the nominal barrier-reduction association also does not persist in ordinal sensitivity analysis. The findings reveal broadly favourable marginal evaluations without a reliably established individual pathway from perceived benefits to willingness. The contribution is to distinguish informational convenience, prospective consideration and justice outcomes, while showing how similar averages conceal different respondent configurations. Implications favour task-specific information support, verifiable sources and human referral, evaluated through comprehension and completed service pathways. The results concern perceptions within a legally oriented sample, not demonstrated improvements in legal assistance, procedural fairness or dispute resolution.
No takes yet. Share an insight, caveat, or question.
Mannan et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: