Agentic Payment Protocols & Agent‑Powered Payments examines the emerging stack that lets autonomous AI agents transact on behalf of users and organizations. Since payment incumbents (for example Visa Intelligent Commerce in 2025) began building agent‑aware rails, attention has shifted to how agent discovery, authorization, settlement and governance interoperate across centralized platforms and decentralized infrastructure such as Etherlink (x402) and rival protocols. This topic covers four intersecting areas: AI agent marketplaces (where agents find services and payment endpoints), agent frameworks (Lindy, IBM watsonx Assistant, Yellow.ai) that create and orchestrate autonomous agents, AI tool marketplaces that expose billable capabilities, and decentralized AI infrastructure that provides on‑chain settlement, identity and auditability. Key tools illustrate the landscape: Visa Intelligent Commerce focuses on secure agent purchases and commerce flows; Lindy targets no‑/low‑code creation and governance of autonomous agents; Yellow.ai and IBM watsonx Assistant address enterprise CX/EX and multi‑agent orchestration; Observe.AI brings voice and conversation intelligence for contact‑center agents and real‑time compliance. Relevance in 2026 stems from growing enterprise deployments of agentic automation, the launch of agent‑aware payment rails, and the push for interoperable standards (e.g., Etherlink’s x402 and peers) that balance user consent, liability, fraud prevention and audit trails. Practical concerns include agent identity and keys, consented micro‑payments, dispute resolution, regulatory compliance, and marketplace governance. Understanding the technical and commercial tradeoffs in these protocols is essential for builders, payments providers and enterprises adopting agent‑powered commerce.