Consumers Prefer Collaborative AI Over Autonomous Shopping

PYMNTS Intelligence's May 2026 Consumer AI Benchmark study, published as "The New AI Handshake: Data Shows When Consumers Want Help and When They Want Control," finds that consumers are adopting AI-enabled commerce features while resisting fully autonomous decision-making in higher-stakes contexts. PYMNTS reports that product discovery, price and deal comparison, and information gathering emerged as natural fits for AI, whereas payments, long-term financial commitments, and irreversible purchases more often prompt consumers to require human oversight.

PYMNTS frames this as evidence that acceptance of agentic commerce will be selective rather than universal: consumers have spent years accepting algorithmic curation (Netflix recommendations, Amazon rankings), but autonomous execution, letting software independently select and purchase products or move money, introduces a different psychological threshold around accountability and control.

Building agentic commerce features should assume mixed adoption by task type rather than blanket acceptance. Interfaces that pair AI recommendations with clear confirmation steps, explanations, and reversible actions are more likely to earn trust in sensitive domains than fully autonomous flows, and teams should instrument consent rates, handoff frequency, and post-decision reversals as early product-viability signals.

Product experiments that surface confirmation checkpoints for financial actions, A/B tests measuring conversion tradeoffs between autonomy and control, and regulatory signals around consumer protections for automated purchasing are the concrete indicators to track.