Agent-based commerce refers to systems where autonomous AI agents—often orchestrated with human oversight—initiate, negotiate, and complete commercial transactions across payments networks and marketplaces. This topic examines platforms and tools that support secure, interoperable transactions (including integrations with Visa, Mastercard and marketplace providers such as Amazon) and the surrounding infrastructure for governance, observability, and regulatory compliance. As of April 2026, adoption is driven by mature APIs from card networks and marketplaces, wider use of tokenization and standardized messaging (ISO/real‑time rails), plus enterprise demand for no‑code/low‑code agent builders and stronger security controls. Key capabilities include multi‑agent orchestration, end‑to‑end audit trails, human-in-the-loop escalation, secure key/token handling, and policy-driven governance to meet PCI, PSD2/open banking, and regional data‑protection requirements. Representative tools: Crescendo.ai combines agentic AI with managed human experts for customer/contact‑center outcomes and dispute resolution; IBM watsonx Assistant enables enterprise virtual agents and multi‑agent automations; Yellow.ai focuses on omnichannel CX/EX automation; Xilos offers infrastructure for visibility into agent activity and connected services; StackAI and Lindy provide no‑code/low‑code agent builders with deployment and governance features; Claude supplies conversational and developer LLM capabilities; Adept’s ACT‑1 specializes in automating multi‑step software workflows. Purchasers should evaluate interoperability with payments APIs, observability and tamper‑evident logging, human failover paths, and built‑in compliance tooling. The market now emphasizes measurable security controls and explicit governance over agent actions as prerequisites for production deployment in regulated commerce environments.