Artificial intelligence is redefining how we buy online. Known as agentic commerce, AI agents can autonomously research, compare, negotiate and complete purchases on behalf of consumers and businesses. The technology relies on structured product data and standardised protocols that allow the agents to discover items, verify authorisation and execute payments securely.
In practice, a user sets constraints—such as budget, product type or delivery time—and the agent handles the entire workflow. The process moves from intent capture to discovery and selection, authorisation via cryptographic proof, and finally payment using secure tokens. The merchant then fulfils the order during checkout and settlement.
Agentic commerce offers faster, cheaper and more personalised shopping experiences. It can be applied to recurring purchases, B2B procurement, travel bookings, ticketing, digital services and physical retail integrations. Yet the shift raises the classic principal‑agent problem in a digital context: can consumers trust the AI to act in their best interests? Issues of alignment, transparency, consent and oversight become critical as agents may optimise for speed or commission rather than user preference.
Other challenges include privacy, data security, accountability, fraud, and the need for interoperable standards such as MCP, ACP and UCP. Despite these hurdles, the trend is growing. IBM reports that 45% of consumers already use AI in parts of their buying journey, and experts predict AI agents could mediate $3‑$5 trillion in transactions by 2027.
As the industry evolves, regulators and businesses must develop frameworks that ensure AI agents act transparently, protect consumer data, and align with user intent, lest the convenience of autonomous shopping be outweighed by hidden costs and mistrust.
