Ahnaf Prio describes agentic commerceAgentic commerce lets AI agents discover, recommend, and purchase goods or services within delegated financial controls. as AI assistance across product discovery, decision-making, checkout, loyalty and fulfilment. Early browser-driving agents were slow, brittle and likely to trigger merchant fraud controls, so newer systems expose structured commerce capabilities rather than imitating a shopper through screenshots and form filling.
Ahnaf Prio maps the emerging protocol stack: MCPModel Context Protocol is a standard way for AI applications to connect with external tools and data sources through consistent interfaces. exposes tools, A2AThe Agent2Agent Protocol is an open communication standard that lets independent AI agents discover capabilities, exchange messages and coordinate tasks. lets customer and merchant agents communicate, ACP and UCP define competing commerce primitives, and AP2The Agent Payments Protocol is an open protocol for expressing and verifying a user's authorization when an AI agent initiates a payment. expresses payment authorization such as spending limits and revocation. Product feeds are sent to large assistant providers in advance so they can index merchant catalogues instead of querying every merchant for every search.
A live bakery-agent demonstration walks through product search, cart creation, checkout state transitions and payment authorization under both UCP and ACP. Ahnaf Prio closes with the operational requirement that retailers build behavioural, protocol-compliance and latency evals, because conversational shopping agentsAn AI shopping agent is a software agent that helps a shopper find, compare and purchase products by communicating with commerce services and merchant systems. can otherwise leak discounts, answer unrelated requests or fail silently as protocols evolve.
Watch on YouTube
