Supertab reposted this
The Rise of Agentic Payment Infrastructure 💡 The goal is simple to describe, hard to achieve: agents that transact autonomously — verifying, approving, and settling payments in real time. Today’s reality: most agentic workflows break down when they reach the transaction layer. Payments are built for human inputs — passwords, card details, manual approvals. The fintech rails are not agent-native. So what do we need to change? 🔹 Identity – Agents must be verifiable: who is acting? What is this AI agent’s purpose? 🔹 Permissions – Clear spending limits and rules of engagement 🔹 Wallets – Programmable funding sources with built-in controls 🔹 Payment rails – Real-time, API-triggerable, built for agents not humans 🔹 Compliance & trust infra – Know Your Agent (KYA), audit logs, risk checks that regulators accept. A big piece of this will be KYA and how to authenticate agents. For agents to transact autonomously, counterparties need to know where the agent comes from, who it is acting on behalf of, and what it is allowed to do. One way to achieve this is through an agent ID or sometimes called a passport - a digital credential that specifies who owns the agent, exactly which agent instance is acting (e.g. a travel bot vs a procurement bot), and what actions it is allowed to perform (such as spending limits, approved merchants, or data access). Until KYA systems like this are widely adopted, most agentic payments will still require a human to step in at critical points. 🔹 Metering & billing – usage-based pricing and outcome-based payments Agentic payments are not only about agents paying online. They are also about agents being paid. If an AI travel agent books a flight for you, or an AI marketing agent optimises campaigns, those services themselves need to be metered, billed, and compensated. This flips the perspective: payments infra must support both sides — agents acting as buyers and agents acting as sellers of services. That means two intertwined problems: 🔸 Agent spending – how an AI reliably executes payments on behalf of a human or company (discussed above) 🔸 Agent earning – how the AI gets credited, invoiced, or rewarded for outcomes delivered. The second piece opens the door to agent metering and billing infrastructure. Instead of flat SaaS fees, we expect usage-based or outcome-based pricing (e.g. “€0.10 per lead validated,” or “1% of ad spend optimised”). We believe this will give rise to a new generation of agent-native companies building metering, billing, and compensation systems. Source: Laura Salesse x Eight Roads - https://shorturl.at/imAEL #Innovation #Fintech #Banking #OpenBanking #API #FinancialServices #Payments #Cards #AI #GenAI #LLM #KYA #Agentic #Agents