Riskified, a New York Stock Exchange-listed company specializing in e-commerce fraud prevention, announced on Wednesday that it has partnered with cybersecurity firm HUMAN Security to develop a unified security framework aimed at protecting merchants in the emerging era of AI-driven shopping agents.
The rise of consumers using large language models — including ChatGPT, Claude, Gemini, Grok, Llama and Perplexity — to research products and compare prices is creating both opportunities and risks for online sellers. While such technology could bring new customers and boost sales, traditional fraud detection systems may fail to recognize fraudulent transactions when they are conducted by AI agents, potentially leading to revenue loss, inventory manipulation and reputational harm.
Riskified said data from its merchant network indicates traffic referred by large language models can be significantly riskier than conventional sources. For example, one ticketing merchant saw LLM-referred traffic that was 2.3 times more likely to be fraudulent than Google search traffic, while an electronics merchant recorded a 1.8-times higher risk rate. The company has also detected signs of automated reseller arbitrage, in which AI agents rapidly purchase products to resell at inflated prices through fraudulent storefronts, which may then be recommended by other AI agents.
To address these threats, Riskified is rolling out new tools, including AI Agent Approve—MCP Server Package on AWS Marketplace for supply-side (merchants) and demand-side (LLMs) to communicate with the Riskified platform APIs and accelerate the safe adoption of AI shopping agents; AI Agent Intelligence, a dashboard to monitor orders from AI agents; and AI Agent Policy Builder, which helps detect policy abuses such as programmatic returns and reseller arbitrage.
Get the Ynetnews app on your smartphone: Google Play: https://bit.ly/4eJ37pE | Apple App Store: https://bit.ly/3ZL7iNv
The partnership will also integrate HUMAN’s newly launched Sightline platform with Riskified’s expertise in fraud prevention and chargeback protection. The companies say this will allow merchants to apply consistent trust policies across human and AI-driven transactions.
“In a world where AI agents transact on behalf of individuals, resolving identity and trust becomes more complex,” said Assaf Feldman, Riskified’s chief technology officer and co-founder. “By working with HUMAN and developing new agentic tools and capabilities, we give merchants a way to safely embrace this shift, turning what could be a threat into a new, profitable digital channel.”
John Searby, HUMAN’s chief strategy officer, said the collaboration would “help establish a trusted ecosystem for agentic commerce” by enabling merchants to approve more legitimate AI-driven orders while blocking sophisticated fraud.