US-based dispute resolution company Chargebacks911 has issued a warning that the payments industry is deploying agentic commerce infrastructure without addressing the post-transaction dispute frameworks needed to support it. The company argues that as Visa, Mastercard, and American Express activate AI agent payment programmes, the infrastructure required to manage disputes, assign liability, and resolve contested charges in agent-initiated transactions remains largely unaddressed.
The warning follows a series of industry developments in early 2026. Visa expanded its Agentic Ready programme globally in late April 2026. Mastercard and Santander completed what they described as Europe's first live end-to-end AI agent payment within a regulated banking framework in March 2026. American Express introduced an agentic commerce developer kit alongside a commitment to cover erroneous purchases made by registered agents on its network.
Structural differences in agentic disputes
Chargebacks911 argues that disputes arising from agentic transactions are structurally different from those that existing chargeback frameworks were built to handle. Traditional dispute resolution rests on the assumption that a human being made a purchase decision, with intent, authorisation, and liability all determined by reference to what a cardholder chose to do. When an AI agent makes a purchase autonomously, acting on delegated authority within parameters set at a prior point in time and without explicit confirmation at the moment of transaction, that assumption no longer applies.
The company's chief technology officer, Donald Kossmann, noted that fraud and dispute systems had been trained over the years to read human behavioural signals, including purchase patterns, device history, and session data. AI agent behaviour is consistent and methodical, disrupting these markers and making it difficult both to defend legitimate transactions and to identify genuinely disputed ones.
Recommended steps and platform response
Chargebacks911 recommends that merchants, acquirers, and financial institutions take three steps ahead of broader agentic commerce adoption. First, establishing granular permission and scope frameworks for AI agents transacting on their platforms, documenting authorisation at the point of delegation rather than reconstructing it after a dispute. Second, building an evidence capture infrastructure that logs agent behaviour continuously across the transaction journey rather than only at checkout. Third, reviewing existing fraud detection rules and dispute thresholds to account for the behavioural differences between human and agent-initiated transactions, as systems calibrated for human behaviour will generate increasing numbers of false positives as agent-led commerce scales.
The company addresses these requirements through its Unified Dispute Management System (UDMS), which uses AI and machine learning to capture consent and permission trails, including what an agent was authorised to do, the scope of that authorisation, and a timestamped record of each action taken. Chargebacks911 supports clients in 87 countries and processes more than 2.4 billion transactions per year through its UDMS and ResolveLab platforms.
