Consumer Acceptance of High-Autonomy AI Assistants Is Driven by Perceived Benefits in Online Shopping Settings Characterized by Scarcity

The Discussion (§10.3; page unavailable) says experiential benefits, not functional benefits, are the central driver of high-autonomy AI adoption under scarcity. The Main Analyses (§10.2.2; page unavailable) likewise report a significant indirect effect through experiential benefits but not functional benefits. The paper does not establish the claimed quality/usefulness → attitude/satisfaction → intention chain. (LLM, conf=0.89)

Checked by a reasoning model — may be wrong.

Section 2 (page unavailable) reports that high AI autonomy reduced online shoppers’ adoption intentions through a diminished personal sense of power. But it identifies this as a finding of Frank and Otterbring (2024), not a demonstration by the cited 2026 paper, and notes that the mediation was not significant under scarcity. The unqualified claim omits that condition. (LLM, conf=0.84)

Study 1 Results (§6.2; page unavailable) report higher click-through for high-autonomy than low-autonomy AI ads (6.02% versus 3.41%). But Study 1 did not manipulate scarcity; the scarcity experiments examined preferences for AI assistants rather than click-through rates. The paper therefore does not establish heightened click-through rates specifically under scarcity-autonomy framing. (LLM, conf=0.88)

The Results (section 6.2; page unavailable) use click-through rate as a behavioral outcome, and the Method (section 6.1; page unavailable) describes the ad campaign as a realistic test prioritizing ecological validity. The paper does not establish that behavioral outcomes are less common, or address conversion, repurchase, or GMV as outcomes; it also cautions in Limitations (section 11.3; page unavailable) that click-through rates may not capture long-term behavior. (LLM, conf=0.86)

The Methods for Studies 2A and 2B describe imagined shopping scenarios in which an AI could make purchases, not observed orders or payments. Study 1 measured ad clicks, and Limitations and Future Research (section 11.3; page unavailable) calls for future studies that track actual purchase behavior. This supports the distinction between simulated execution and observed transactions in this paper, but does not establish that observed agentic transactions are rare across the wider literature. (LLM, conf=0.84)

The Limitations and Future Research section (11.3; page unavailable) explicitly says the studies were confined to online shopping and a narrow set of utilitarian electronic products, supporting the limited-topic example. The Method section (7.1; page unavailable) says participants were recruited through Prolific, but the copy does not establish that panels are often nonrandom samples. It also does not establish a general pattern of studies limited to one platform or a few task types. (LLM, conf=0.86)

Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

The Theoretical Contribution (§11.1) says consumers are more willing to delegate decisions to high-autonomy AI under scarcity, while the Practical Implications (§11.2) say such assistants can be expected to underperform in routine shopping. The paper discusses transparency safeguards, but it does not establish that education, safeguards, and trust recalibration are needed for autonomy effects to extend to routine shopping. (LLM, conf=0.86)

The Method (section 9.1) measures adoption through self-reported intentions, and Limitations and Future Research (section 11.3) says the study relies on short-term proxies and calls for research tracking actual purchases. The paper does not establish that this is true of many studies, and it does not report self-reported ordering as an outcome; its field experiment also measured actual ad clicks (section 6.2). (LLM, conf=0.84)

The Results (§6.2) report higher ad click-through for the high-autonomy AI assistant, and the Limitations (§11.3) say the adoption benefit is confined to scarcity settings, supporting a positive but non-universal response. The paper does not establish observed conversions or purchases caused by reduced friction or better-fitting recommendations; §11.3 explicitly calls for future studies tracking actual purchase behavior. (LLM, conf=0.86)

Section 4 suggests that fast, efficient decisions can make high-autonomy AI appealing in scarcity-related shopping, and Section 2 reports that preference-matched recommendations are associated with greater satisfaction than mismatched ones. But the paper does not show behavioral lift from reduced latency or fewer steps, or from improved recommendation relevance. Its own results (Section 10.2.2) identify experiential benefits—not functional benefits—as the significant pathway to adoption under scarcity. (LLM, conf=0.83)

The Main Analyses (§9.2.2; page unavailable) find a positive indirect effect on AI adoption intentions through perceived benefits under scarcity, but not under regular shopping conditions. The Abstract (page unavailable) reports lower willingness to use a high-autonomy assistant than a low-autonomy one. The paper does not establish trust as the pathway, and its context-dependent findings do not support calling the effect mostly positive. (LLM, conf=0.88)