Agents of Change: How Agentic AI Is Redefining Commerce

From early chatbots to newer generative models, artificial intelligence (AI) has been a key disrupter of the digital economy. However, these models still rely on human input. Now a new iteration points the way toward the future of eCommerce: agentic AI, an autonomous actor capable of moving value through this system with unprecedented speed.

AI-powered digital agents are more than just intelligent shopping assistants. They are independent proxies, acting without assistance to search, compare and transact based on a human shopper’s pre-programmed preferences and imperatives. By enabling personalized, frictionless interactions and reducing consumer decision-making burdens, agentic AI is expected to revolutionize commerce and create new ways for merchants, banks and technology providers to interact with customers. Nevertheless, this will be possible only when the technology overcomes the significant trust and compliance hurdles standing in its way.

• From Assistant to Agent: Defining AI’s Next Growth Phase in Commerce
• Fears and Frictions: Challenges to Widespread Adoption
• From Risk to Resilience: Building Agentic AI for Scale
• Seizing the Agentic Commerce Advantage

From Assistant to Agent: Defining AI’s Next Growth Phase in Commerce

Agentic AI represents a transformative shift for global commerce, promising greater convenience for consumers and new efficiencies for businesses. With nearly half of consumers open to using the technology, major payments and FinTech firms are stepping up innovation to seize the moment.

Agentic AI can provide a more frictionless consumer shopping experience.

Many consumers are ready for this shift. In fact, nearly half (45%) would be comfortable allowing AI agents to complete purchases on their behalf. Among Gen Z consumers, that figure rises to a majority share (54%).

Many retailers are gearing up for this shift—and those that aren’t risk falling behind.

This transformation is already underway, with AI innovators, including OpenAI, Perplexity, Google and Microsoft, as well as Amazon, rolling out features in which autonomous agents can complete shoppers’ orders. As such, merchants are looking to adapt to make their products more findable by AI systems.

Retailers are also looking for ways to use agentic AI on their own platforms. Specifically, 43% are piloting autonomous AI. Confidence is high: 81% of retailers trust AI’s ability to operate autonomously, provided the right guardrails are in place. The shift is coming whether retailers are ready or not, with the agentic commerce market projected to reach $1.7 trillion by 2030.

Fears and Frictions: Challenges to Widespread Adoption

Fears and frictions temper agentic commerce’s promise. Fraud and compliance issues pose significant hurdles. With merchants bracing for more AI-driven attacks and the vast majority of consumers expressing concerns, building trust and navigating global regulations are critical for agentic commerce to reach scale.

Nearly all consumers have their doubts about agentic commerce.

Top worries about agentic AI vary by region. According to a newly published Worldpay report, individuals in the United Kingdom and China are most worried about incorrect purchase decisions, while consumers in the United States, Australia and France are most nervous about identity theft. Those in Brazil cite fraud as their leading concern, and consumers in Singapore are most apprehensive about unauthorized purchases.

These fears are compounded by broader doubts about AI’s reliability. Past experiences with AI have shaken consumers’ trust. Nearly all internet users (96%) are aware of AI’s tendency to hallucinate, and 86% have personally encountered the phenomenon.

Merchants are concerned as well, not just about fraud but also about liability if things go wrong.

Roughly two-thirds of merchants (65%) report being aware of AI-driven fraud attempts targeting them or their customers in the past year. Nearly all (99%) express at least some concern about AI-related fraud, with 46% saying they are very concerned. For instance, malicious AI agents could mimic consumers or exploit weak defenses.

The increasing use of agentic AI in commerce also raises concerns about liability when AI agents make errors, such as incorrect purchases. Ambiguities exist regarding responsibility for issues like fraud and chargebacks, highlighting the need for industry standards before widespread agentic AI adoption. Potential liable parties include the consumer, the merchant or the AI platform.

Regulatory scrutiny presents another hurdle to agentic commerce’s growth.

The complexity of complying with each market’s unique privacy regulations is slowing agentic AI’s global expansion. Experts note that most firms seem unprepared to navigate these disparate policies. Evolving laws could cause some systems to become obsolete in the future. A California Management Review article recommends that organizations develop governance frameworks to manage the ethical and operational risks of agentic AI. These frameworks should emphasize establishing oversight, promoting transparency, embedding ethical checks and prioritizing security and data protection in the use of the technology.

From Risk to Resilience: Building Agentic AI for Scale

Industry leaders are tackling agentic commerce’s trust challenges through innovation, collaboration and consumer education. New solutions, partnerships and protocols exemplify efforts to strengthen identity verification, prevent fraud and build consumer trust through transparent and secure AI-driven payment systems.

New tech solutions address key challenges to laying the groundwork for agentic commerce’s expansion.

AI leaders are collaborating with specialists to establish this safety and trust.

Google recently announced the launch of its Agent Payments Protocol (AP2), an open protocol designed in line with industry rules to enable secure payments of all kinds via AI. The protocol was developed in collaboration with more than 60 organizations, including payment firms such as Worldpay as well as marketplaces and merchants. The move reflects the broader industry consensus that success depends on robust verification, strong fraud defenses and transparent data practices.

Seizing the Agentic Commerce Advantage

Agentic AI-driven retail is no longer theoretical—it’s here and already beginning to transform commerce. Businesses that act now will set the benchmarks for trust, speed and customer experience. Those that delay risk being eclipsed by early adopters defining the next digital economy.

PYMNTS Intelligence offers the following roadmap for firms preparing to lead in agentic commerce:

• Build trust as the cornerstone. Many consumers are open to agentic AI but have lingering worries about the technology. Addressing concerns around fraud, transparency and data use is essential to winning confidence.
• Invest in resilient infrastructure. As fraud threats and regulatory demands intensify, partnerships with identity verification, payments and fraud-prevention specialists are key to scaling securely.
• Embed agentic AI into core commerce strategies. From autonomous shopping agents to intelligent product discovery, integration must move beyond experimentation. Businesses that embed agentic AI across customer touchpoints can simplify operations and unlock new efficiencies.
• Shape industry standards through early adoption. By deploying agentic AI now, firms can influence emerging protocols and best practices—such as secure agent-led payment frameworks—ensuring alignment with both regulators and customers.

By taking these steps, businesses can move agentic commerce beyond experimentation and turn it into a competitive advantage. The shift is already in motion, and those that act now will not only capture market share but also help shape the standards of the next digital economy.

The story of agentic commerce is ultimately a story about confidence—confidence in technology, in partnerships and in possibility. As AI agents begin to transact for us, our collective responsibility is to ensure they do so safely, transparently and for the benefit of everyone.”