Smart Shopping Agents Still Need Someone to Vouch for Them

AI summary of the source article

Speaking to PYMNTS, Synchrony SVP Mike Storiale explained that moving from AI product recommendations to autonomous transactions requires significant foundational infrastructure. This includes common commerce protocols, authentication, proof of intent, and a rebuilt trust layer capable of verifying delegated consumer authority. Storiale noted that AI agents could shift financing decisions upstream into the product discovery phase, rather than leaving them at checkout. To test this, Synchrony has utilized a ChatGPT plugin to display financing terms and deals. Storiale emphasized that adoption cycles for payments standards, risk controls, and merchant integrations typically take years, making infrastructure and trust the true bottlenecks for machine-initiated transactions.

Why it matters

For financial institutions and card issuers, AI agents could determine purchase recommendations based on early financing options, diminishing the importance of traditional checkout placement.

Key facts

• Synchrony is testing agentic financing options through a ChatGPT plugin that surfaces terms, offers, and deals within the AI environment.
• Critical infrastructure requirements for machine-initiated purchases include proof of intent, authentication, common commerce protocols, and delegated authority controls.
• Storiale identifies adoption speed as a major overhyped assumption, noting payments standards, integrations, and risk controls take years to scale.