PSE Consulting research has found that consumers prefer free, ad-supported AI shopping tools over paid impartial alternatives, though trust concerns remain.
The study, which surveyed 4.250 consumers across the UK, the US, France, and Germany, found that 43% of respondents would choose a free AI shopping assistant whose recommendations are influenced by advertising, compared with 27% who would pay for a fully impartial alternative. The findings indicate that consumers are carrying established digital trade-offs (most notably the exchange of advertising exposure for free services) into the emerging context of agentic AI commerce.
Trust sensitivity and market variation
Despite this general openness, the research identifies a significant constraint. Four in ten consumers (40%) say advertising would reduce their trust in AI-generated recommendations, with that figure rising to 48% among UK respondents. At the same time, trust sensitivity is notably higher among older consumers: half of those aged 55 and over say advertising would reduce their trust, compared with a third of under-35s. The latter group, having grown up with algorithmically curated services, applies a different baseline expectation of what neutral recommendation looks like.
The US stands out as the strongest market for paid alternatives, with 34% of US consumers willing to pay for a fully impartial AI assistant, representing the highest proportion among the four markets surveyed.
The research does not find evidence of widespread uncritical deference to AI decision-making at this stage. Instead, consumers appear to retain awareness of commercial influence within AI shopping environments, though the study notes these behaviours remain characteristic of relatively early adopters of AI-assisted commerce.
Implications for the AI commerce ecosystem
In addition, the findings point to a likely tiered commercial structure for AI shopping tools. Free, ad-supported models are expected to dominate mainstream adoption, with seller-funded placements and sponsored recommendations becoming normalised features. Alongside these, premium agnostic models and enterprise services may develop for use cases where user-centric advice is a primary requirement.
Chris Jones, Managing Director at PSE Consulting, noted that the data reflects a more complex monetisation landscape than a binary choice between ad-supported and paid models, adding that the real risk is a pattern of gradual degradation in user experience ( representing a dynamic the firm associated with the concept of platform 'enshittification') if commercial influence is not carefully managed in AI tool design.
Beyond the consumer-facing layer, the research highlights a distinct challenge for merchants and platforms. As agentic AI models become the primary discovery interface, enterprises are expected to invest in ensuring their products are accurately represented, correctly prioritised, and consistently interpreted within AI-driven recommendation environments. In this context, trust becomes both a consumer experience and something engineered at the infrastructure level.
PSE Consulting's study forms part of a broader research programme examining the commercial and structural implications of agentic commerce across four major Western ecommerce markets, with a full report forthcoming.
