The Wide Gap Between AI Bot 'Shopping' and 'Payments' — Where Agentic Commerce Stands Today

Bot Payments Lag Behind in Agentic Commerce

Bot payments lag in agentic commerce

Within the emerging world of agentic commerce, a broad gap exists between bot shopping and autonomous payments.

On January 29, 2026, Payments Dive reported that a significant gap exists between "product search and comparison" and "autonomous payment completion" in agentic commerce — the model where AI agents shop on behalf of consumers. While AI bots excel at product research and comparison, autonomous payment completion remains distant due to challenges in authorization, compliance, and payment infrastructure.

The article suggests that AI shopping bots will initially handle everyday items such as milk, toothpaste, and cat litter. By starting with low-price, low-risk, and low-complexity products, consumers will go through a familiarization phase as they grow accustomed to this new purchasing experience.

Industry Landscape

Agentic commerce has rapidly become a central industry theme since the start of 2026. Consulting firm Flagship Advisory Partners stated in a January 21 market report that this bifurcation between shopping and payments will persist "for the time being." Analysts Ben Brown and Pavle Stamenic noted that "agentic payments depend on industry progress in building trust, governance, and interoperability."

The scale of the market is also becoming clearer. According to a Boston Consulting Group study, approximately 81% of consumers are open to using agentic commerce tools, potentially influencing $1.3 trillion in spending. Additionally, a McKinsey report from October 2025 estimates up to $1 trillion in US retail sales opportunity by 2030, and $5 trillion globally.

Meanwhile, an explosion of competing protocols has emerged ahead of this market's rapid growth. Visa and Mastercard are each developing proprietary agentic commerce protocols. Google announced the Universal Commerce Protocol (UCP) at the January 2026 NRF (National Retail Federation) annual conference, partnering with over 20 companies including PayPal and Walmart. Additionally, another protocol led by OpenAI and Stripe also exists.

Why Bot Payments Are So Difficult

The payment lag in agentic commerce has structural reasons.

Authentication and authorization challenges: Traditional payment infrastructure was designed with human consumers in mind. As the McKinsey report points out, when the "customer" is an AI agent, "new approaches" are needed for consumer authorization, programmed spending limits, and consent management. Jim Nguyen, CEO of agentic payment startup InFlow, stated that "an entirely new mental model is needed — one that treats agents as independent customers tied to their actual owners."

Fraud prevention challenges: Existing fraud detection systems flag unusual purchase times and rapid successive purchases. However, AI agents exhibit exactly these behavioral patterns, making it difficult to distinguish legitimate agent transactions from fraudulent bots. Visa addressed this by introducing the Trusted Agent Protocol in October 2025 with over 10 partners, building an open framework to differentiate malicious bots from legitimate AI agents.

Liability attribution: Traditional payment disputes have been processed among four parties (consumer, merchant, issuing bank, and network), but with the addition of AI platforms as a fifth player, liability attribution becomes unclear when an agent makes an erroneous order.

Parijat Banerjee of consulting firm LatentView Analytics projects that fully autonomous payments will not become mainstream until "late 2027 to 2028."

Impact and Applications for E-Commerce Operators

Higher conversion rates: Chris Jones of PSE Consulting, citing Adobe Analytics and Klarna data, notes that agentic commerce channels deliver "conversion rates approximately 30% higher than other channels." This figure provides e-commerce operators with a compelling case for investing in agent support.

Phased adoption is realistic: John Lunn, CEO of payment orchestration startup Gr4vy, states that even small merchants can manage and experiment with agentic transactions as an "independent bucket" while leveraging existing payment infrastructure. Rules such as "no agent orders over $100" or "no luxury brand items sold via agents" are configurable.

Choose protocols carefully: Adam Behrens, CEO of software startup New Generation, warns that "if you invest resources into one ecosystem or platform now, you'll have to start over when the market shifts in three or six months." Gr4vy also predicts that half of the protocol implementations currently underway will be discarded by year-end.

Watch eBay's moves: eBay has banned unauthorized AI bots from transacting on its marketplace effective February 20, 2026. This serves both as a strategy to build its own agent experience and as a defense against revenue erosion from unauthorized bots. E-commerce operators similarly need to proactively decide which agents to grant access to their inventory.

Conclusion

Agentic commerce is definitively underway as the next major evolution in e-commerce. However, the gap between "search and comparison" and "payment completion" will not be easily bridged, either technically or institutionally. 2026 is a transitional year where major players like Visa, Mastercard, and Google compete over protocols and industry standards take shape.

What matters for e-commerce operators is adopting an approach of "experimentally" introducing agentic transactions separated from existing infrastructure, without over-committing to any specific protocol. As the 30% conversion rate improvement suggests, the potential is significant, but preparations for fraud prevention, liability, and inventory management must advance simultaneously. Fully autonomous agent payments are not expected to become mainstream until late 2027 or beyond — now is the phase to "learn while preparing."