Merchants’ fear of AI shopping agents is just a ‘hallucination’, four-wave international study finds. Based on responses from 4,250 consumers across the UK, US, France and Germany, the research explores the emerging ‘shortlist economy’ and how AI is reshaping the path to purchase.
While AI is increasingly influencing which products consumers consider, it isn’t replacing the brands and marketplaces they ultimately choose to buy from.
Among the key findings:
- 89% say recognising the seller’s brand is important or very important when acting on an AI recommendation, and 93% say the same of customer reviews.
- Just 14% of consumers say they simply follow an AI assistant’s top recommendation, fewer than half the 32% who cite price as their primary decision driver.
- 93% expect their use of retail marketplaces to stay the same or increase as AI adoption grows; 89% say the same for travel and for food delivery platforms.
- 68% would consider an AI’s pick of an unfamiliar brand, but only after checking reviews and ratings first.
- 74% prefer an independent AI assistant over one embedded in a single retail platform: 41% want a universal tool such as ChatGPT or Gemini, 33% want a vertical specialist (travel, finance, healthcare), and just 10% prefer a platform-embedded assistant.
- Only 3% describe their experience with agentic commerce as negative so far — 43% call it mostly positive, 43% mixed. Germany is the most cautious market (34% mostly positive versus 52-53% in the US and France).
“The fear that AI agents will simply cut merchants out of the transaction is, on the evidence of our research, a hallucination,” says Chris Jones, Managing Director at PSE Consulting.
The stakes extend beyond consumer sentiment. PSE’s own analysis estimates that around $1.25 trillion in e-commerce spend could migrate to agentic rails by 2030, with travel and accommodation expected to account for a significant share. That scale is why Visa, Mastercard, Stripe and Google are already building infrastructure and standards for agentic payments, rather than waiting for the model to mature.
The open question is how agents, merchants and PSPs coordinate without duplicating each other’s roles. PSE points to Google’s Universal Commerce Protocol (UCP) as an early attempt to let agents retain visibility into consumer choice and transaction execution while merchants keep control of stock, payments and fulfilment, and PSPs orchestrate authentication, fraud and chargeback processes across channels. Consistent data-sharing standards between all three layers, PSE argues, are a prerequisite for agentic commerce to scale without creating new liability gaps.
Without that intra-industry coordination, disputes like the one between Amazon and Perplexity, where the e-commerce platform accused the AI developer’s Comet AI browser of accessing Amazon’s password-protected accounts, including Prime subscriptions, without Amazon’s authorization, may continue to arise from time to time. That might have been the first real test of who counts as an “authorized” visitor when the visitor is an AI agent. Yet, PaySpace Magazine Global team believes it may not be the only one, until general rules and support rails are established for the emerging agentic economy. Amazon itself, that has been so fervent about AI bots violating its terms of service, has since been at the core of a similar scandal when small retailers discovered their products listed and sold on Amazon without their consent via the platform’s new AI agent technology.
The concluding PSE report, The Shortlist Economy — How AI Is Rewiring Buying Habits (and How Merchants Can Respond), is available to access here.
