J.P. Morgan Payments and Mirakl Enable AI Agents to Buy Autonomously
"We are entering an era where AI agents won't just assist with shopping, they will transact"
J.P. Morgan Payments and Mirakl announced a strategic global partnership on March 10, 2026, designed to power agentic commerce at enterprise scale.
The collaboration pairs Mirakl's commerce platform with J.P. Morgan's payment infrastructure to enable AI agents to autonomously discover products, evaluate options, and complete transactions on behalf of consumers and businesses.
The partnership addresses a fundamental shift in online shopping. AI agents are moving beyond recommendation and research roles into actual purchasing, creating new demands for both commerce infrastructure and payment security.
Mirakl Nexus will serve as the commerce layer, optimizing product catalogs for AI discovery and managing the complete shopping journey from discovery through checkout and post-sale service.
J.P. Morgan Payments will provide the payment processing, tokenization, and fraud protection required to make those transactions secure and auditable at scale.
"We are entering an era where AI agents won't just assist with shopping, they will transact," said Mike Lozanoff, Global Head of Merchant Services at J.P. Morgan Payments. "As agents move from browsing to buying, the differentiator won't be AI, it will be governance: identity, consent, limits, and interoperability at global scale."
The partnership comes as AI agents are evolving from research assistants into autonomous purchasing systems, with major language models like ChatGPT, Gemini, and Copilot beginning to integrate commerce capabilities.
The Governance Challenge
The shift from AI-assisted browsing to AI-executed purchasing introduces complex operational and risk management questions. When an AI agent completes a transaction without direct human intervention at the point of sale, traditional payment security models built around human verification become insufficient.
The partnership is designed to solve this by embedding verified agent identity, user-controlled permissions, and bank-grade risk management directly into the payment infrastructure.
J.P. Morgan Payments is building tools to help merchants integrate with consumer-facing AI agents, manage agentic fraud risk, and support agent-driven sales channels.
The company is also working with industry stakeholders to define standards for how AI agents authenticate, receive permissions, and execute purchases within defined limits.
Mirakl's role centers on making product information accessible and actionable for AI systems. Mirakl Nexus optimizes catalogs for AI discovery, ensuring agents can access accurate product details, inventory data, and pricing. The platform also enables merchants to sell directly through large language model interfaces including Gemini, Copilot, and Perplexity.
For retailers and merchants, the integrated solution removes the burden of building custom infrastructure to participate in agentic commerce.
Rather than developing separate integrations for commerce orchestration and secure payment processing, merchants can access both through a unified platform.
The partnership also repositions how merchants compete in AI-driven shopping environments. Traditional e-commerce competition has centered on product selection, pricing, and logistics. In agentic commerce, differentiation increasingly depends on how intelligently merchants serve AI agents.
Retailers that provide richer product metadata, cleaner catalog structures, and more sophisticated recommendation capabilities become more discoverable and favorable to AI systems making purchasing decisions on behalf of consumers.
"Mirakl Nexus is key to unlocking agentic commerce, optimizing product catalogs for AI discovery and enabling merchants to sell directly through LLM channels," said Adrien Nussenbaum, Co-founder and Co-CEO of Mirakl. "J.P. Morgan Payments brings the payment and risk management capabilities that enable AI agents to support user-verified purchases securely and at enterprise scale."
The solution is currently in closed beta with select retailers and merchants, with broader availability planned for 2026. The companies did not disclose pricing structures or specific merchant participants in the beta program.