Agentic commerce boom creates new risks for merchants

The rise of agentic commerce is rapidly reshaping the future of online retail, with enterprise merchants moving at speed to ensure their businesses are visible to AI-powered shopping assistants.

Agentic commerce creates new risks for merchants

However, as adoption accelerates, concerns are growing that many organisations may be embracing the technology before fully understanding the fraud, security and liability implications.

According to Ravelin’s Agentic Commerce and Fraud Report 2026, the ecommerce sector is experiencing a surge of interest in autonomous AI agents capable of researching products, comparing options and completing purchases on behalf of consumers.

For many merchants, the technology represents the next major shift in digital commerce, comparable to the emergence of mobile shopping or digital wallets.

The report found that 44% of enterprise merchants have already begun integrating agentic commerce protocols into their operations, while a further 32% plan to do so within six months. Just 6% currently have no plans to adopt the technology. Meanwhile, 42% of merchants are exploring the development of proprietary AI shopping agents to create more direct relationships with customers.

The Commercial Case for Agentic Commerce

Merchants see significant commercial potential in agentic commerce. As consumers increasingly rely on AI assistants to make purchasing decisions, businesses are concerned that failing to integrate with agentic ecosystems could reduce visibility and limit future sales opportunities.

Respondents to the survey expect agentic transactions to account for between 6% and 30% of ecommerce volumes within three years. This anticipated growth is driving investment despite the technology remaining in its early stages.

For payment providers, card issuers and merchants alike, agentic commerce promises more efficient customer acquisition, highly personalised purchasing experiences and potentially higher conversion rates.

Yet the enthusiasm surrounding these benefits is accompanied by a growing recognition that entirely new categories of risk are emerging.

Fraudsters Are Already Eyeing the Opportunity

Historically, fraud has evolved alongside every major innovation in digital commerce, and agentic commerce appears unlikely to break that pattern.

Ravelin highlights a number of emerging threats, including agent hijacking, malicious AI systems impersonating legitimate shopping agents, and the creation of fake environments designed to harvest payment credentials and sensitive customer information.

Criminals could also deploy their own autonomous agents to automate fraudulent activity at scale, dramatically increasing the volume and sophistication of attacks.

Traditional fraud prevention systems, many of which are designed to identify suspicious human behaviour, may struggle to distinguish between legitimate and malicious AI-driven transactions.

The report suggests that businesses must begin developing entirely new approaches to authentication, identity verification and transaction monitoring if they are to manage these risks effectively.

Who Is Liable When Things Go Wrong?

One of the most pressing issues facing the payments industry is the question of liability.

Existing ecommerce frameworks clearly define the responsibilities of merchants, consumers, payment providers and financial institutions. Agentic commerce introduces an additional layer into the transaction chain, creating uncertainty when disputes arise.

If an AI agent completes a purchase using stolen payment credentials, who should bear responsibility? The merchant? The AI provider? The protocol developer? Or the consumer whose agent initiated the transaction?

Similar questions apply to chargebacks, refund fraud and disputed purchases. Until regulatory frameworks and industry standards evolve, merchants could find themselves exposed to significant legal and financial uncertainty.

Trust Gap Remains a Challenge

The research also revealed an intriguing contradiction. More than half of merchants surveyed said they trust AI agents more than human shoppers. Rising levels of refund abuse, account takeover fraud and policy exploitation may help explain this sentiment.

However, consumers remain considerably more cautious. More than half expressed concerns about fraud and security when allowing AI agents to make purchases on their behalf.

As agentic commerce moves from experimentation to mainstream adoption, success is likely to depend on striking the right balance between innovation and risk management.

Businesses that invest in fraud prevention, governance and clear accountability frameworks today will be better positioned to capitalise on what could become one of the most significant developments in ecommerce and payments over the coming decade.