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

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

Summary

This paper examines when and why consumers accept high-autonomy AI shopping assistants (systems that can buy on a user’s behalf). Across five studies (including a field experiment with real ad clicks and four online experiments with nearly 2,000 participants), the authors find that: under ordinary shopping conditions consumers prefer lower-autonomy assistants, but in settings marked by scarcity (e.g. limited stock or time-limited deals) consumers are more likely to choose high-autonomy AI. The effect is robust across samples and manipulations and is driven primarily by perceived benefits—especially experiential benefits such as relief, enjoyment and reduced stress—not functional or symbolic gains.

Method highlights: a behavioural field test (Google Ads; N impressions = 1,102) showed nearly double click-throughs for high-autonomy wording versus low-autonomy. Controlled experiments (US and UK samples) manipulated scarcity (time vs availability) and autonomy level, added a ‘no choice’ option, and used multi-dimensional benefit scales to test mediation.

Key Points

1. Behavioural evidence: ads framing an AI as “auto-purchase” generated higher click-through rates (6.02% vs 3.41%) than “notification” framing.
2. Scarcity is a boundary condition: consumers are significantly more likely to choose high-autonomy AI under scarcity than in regular shopping settings (replicated across studies and scarcity types).
3. Perceived benefits—especially experiential benefits—mediate the effect: in scarce contexts high-autonomy AI is seen as more enjoyable/relieving, which drives adoption.
4. Functional and symbolic benefits did not account for the effect; experiential (hedonic/affective) gains were the unique mediator.
5. Implications for firms and policy: position high-autonomy features during scarcity events (e.g. limited-time sales) and ensure transparency about delegation consequences (payments, returns, retailer choices).

Context and relevance

This research updates the debate on algorithm aversion by showing that acceptance of autonomous decision-making by AI depends on context. It bridges lab-based findings and real-world behaviour, arguing scarcity amplifies the value of delegating decisions to AI. For product teams, marketers and policymakers, it clarifies when to promote or restrain high-autonomy features and flags the need for clearer transparency and regulation (for example in light of the EU AI Act).

Author’s take (punchy)

Big picture: high-autonomy AI isn’t universally scary — people will hand over control when there’s something to lose. The authors ran ads, experiments and mediation tests to show it’s the feel-good payoff (getting the deal, less stress) that wins consumers over in scarce situations. If you’re designing AI shopping services, think about timing and messaging: scarcity + experiential framing = higher uptake.

Why should I read this?

Want the short version? If you build or sell AI shopping tools, this paper tells you when consumers actually welcome an AI that buys for them — spoiler: during flash sales and limited-availability drops. The authors back that up with real ad clicks and clean experiments, so you don’t have to wade through loads of theory to know when to push an “Auto-Purchase” feature or dial it back.

Source

Source: https://onlinelibrary.wiley.com/doi/10.1002/mar.70074?af=R