The benefits & risks
of agentic AI in ecommerce
Introduction
Agentic commerce is a profound transformation in the way we sell and buy. Just as brick-and-mortar stores have been succeeded by the clicks of online shopping, today’s search-based discovery and online stores are giving way to new experiences built around LLMs and AI shopping assistants. McKinsey predicts that agentic commerce in the US retail market alone could orchestrate up to $1 trillion USD in revenue by 2030, rising to $5 trillion globally. That’s a powerful channel for growth.
In this exclusive report from The Paypers, supported by Riskified, we’ll look in depth at what this means for consumers, merchants and the payments industry. We’ll cover the different agentic models, the opportunities and risks, plus what agentic commerce means for fraud and the customer experience. Along the way, you’ll hear from experts in the payments and retail sectors, including a Q&A with Assaf Feldman, Riskified’s co-founder and Chief Strategy Officer – Technology.
Agentic commerce alters the risk landscape, adding unforeseen threat vectors as well as new operational challenges. Through its powerful AI platform, wealth of network data and identity engine, Riskified is poised to help merchants rise to these challenges. Read on to discover what agentic commerce might mean for your ecommerce stores and financial platforms, and how to maximize its opportunities.
Executive summary
Agentic commerce represents a paradigm shift for ecommerce, as we transition from search-based online shopping to an ecosystem driven by large language models (LLMs) and autonomous agents. These agents select products, negotiate deals and can complete transactions on behalf of users, with massive implications for merchants, payment service providers and consumers.
- Ecommerce has evolved dramatically over the last thirty years and now AI is powering an online retail revolution. Using AI models like ChatGPT or Google Gemini for product discovery and comparison is just the start. Agentic shopping and customer service are the next frontier.
- The emerging agentic marketplace is split into two distinct models. Store-native agents provide pre-sales guidance and assistance and, potentially, post-sales support. By maintaining merchant control and a coherent customer experience, they offer a practical and accessible route into agentic commerce. Meanwhile, AI assistants select and may purchase products directly for consumers, comparing multiple merchant sites. They’re limited in their ability to complete transactions, but this is coming. While powerful, merchants must consider the potential impact on customer engagement and brand control.
- When AI agents make autonomous purchase decisions, how can customer intent be verified? Merchants need clear tracking of intent and authorization to avoid operational risks. The payments industry and technology providers are working to fill the intent gap through tokenized frameworks, cryptographically signed mandates, telemetry and tamper-proof audit trails.
- Agentic commerce disrupts existing anti-fraud models, while creating new vulnerabilities – from hijacked agents to credential theft to fraudulent chargebacks and prompt injection steering. New protocols and frameworks are equipping merchants and PSPs with the tools they need to fight agentic fraud. The emergence of specialist agents trained specifically to combat it could prove pivotal.
- Riskified co-founder and Chief Strategy Officer – Technology, Assaf Feldman explains why a ‘wait and see’ approach to agentic commerce won’t work, and how specialized trust agents can operate directly in the workflows to validate identity, interrogate behavioral history and block fraud.
Submit the form to continue reading
Subscribe to our newsletter updates
Thank you for subscribing!
Stay tuned for next month’s edition of RiskiNews for the latest trends in fraud and policy abuse. We’ll deliver it straight to your inbox.