The emergence of AI agents capable of making autonomous purchasing decisions is reshaping the payments landscape. Agentic commerce—where artificial intelligence systems act on behalf of consumers to discover, negotiate, and complete transactions—presents both extraordinary opportunities and complex challenges for the payments ecosystem.
The Promise and the Problem
AI agents are poised to transform commercial transactions. Unlike traditional e-commerce where humans click through checkout flows, agentic commerce enables software to make purchasing decisions based on learned preferences and sophisticated algorithms. An agent might automatically reorder groceries, negotiate insurance rates, or book travel that optimizes for cost and convenience.
For payments providers, this opens remarkable opportunities. Transaction volumes could multiply exponentially as agents handle routine purchases. Micro-transactions become viable without human approval friction. Agents enable dynamic pricing and real-time optimization of payment methods based on rewards or exchange rates.
The fundamental challenge is determining when an AI agent has legitimate authority to spend money. How do we verify that a transaction genuinely reflects the owner’s intent rather than a hallucination or compromise? Current authentication frameworks were designed for human-initiated transactions, making Strong Customer Authentication and similar regulations problematic when software makes dozens of autonomous daily decisions.
Liability and Trust
When an AI agent makes erroneous purchases—ordering 50 pounds of bananas instead of 5—who bears responsibility? The consumer, agent provider, or merchant? Payment networks have chargeback systems for traditional commerce, but these assume human decision-making. New frameworks must distinguish between agent errors, consumer miscommunication, merchant misconduct, and system compromises while balancing consumer protection with practical adoption.
Effective AI agents require extensive data about preferences, habits, and financial constraints. Payment providers will have unprecedented visibility into purchasing patterns and AI decision-making logic, creating significant privacy risks. Questions about data portability and preventing manipulation of agent behavior will shape competitive dynamics.
Infrastructure Evolution
Payments infrastructure must evolve substantially for agentic commerce. Current latencies acceptable for humans become problematic for agents making rapid decisions. Real-time payment rails may become essential rather than premium features. Standardized APIs will be critical for agents to discover payment options and execute transactions across diverse merchants—without them, the ecosystem fragments.
Fraud detection systems currently identify anomalous human behavior patterns. When AI agents routinely make purchases across geographies or categories, these signals become unreliable. Payment providers need new approaches that distinguish legitimate agent behavior from compromised systems, creating a complex dynamic where machine learning monitors other machine learning.
Regulatory Challenges and Opportunities
Regulators are only beginning to address agentic commerce. Existing consumer protection laws assume humans make purchasing decisions with conscious intent. Anti-money laundering and Know Your Customer requirements were designed for people, not software agents. New frameworks must address agent transparency, consent, and accountability without stifling innovation or pushing commerce into unregulated spaces.
Despite these challenges, agentic commerce creates tremendous innovation opportunities. Companies solving authentication could enable new business models. Payment networks adapting fastest to agent-driven patterns could capture disproportionate value. AI could make payments more inclusive by helping consumers optimize choices and navigate complex fee structures.
Building the Future
The transition to agentic commerce will be gradual. Early adoption will focus on low-risk, high-frequency transactions like recurring subscriptions and routine household purchases. As trust builds, agents will handle higher-value and more complex transactions.
Success requires collaboration across payment networks, financial institutions, technology companies, merchants, and regulators to establish standards and build protective systems. The companies viewing agentic commerce as partnership opportunity rather than threat will thrive.
The convergence of AI and payments represents a fundamental shift in commercial infrastructure. While challenges around authentication, liability, privacy, and regulation are substantial, they’re not insurmountable. By approaching these issues thoughtfully and collaboratively, the industry can unlock agentic commerce’s potential while protecting consumers. The question isn’t whether AI agents will reshape payments, but whether we’ll build that future thoughtfully or haphazardly.
Business Innovation by Donna Bailey 