IXOPAY and Zip have jointly announced an open industry framework designed to address trust and liability challenges in AI agent-initiated payments.
The announcement responds to a structural gap in existing payment infrastructure. The traditional four-party model was built on the assumption that a human initiates each transaction. As AI agents begin acting autonomously on behalf of consumers, that assumption no longer holds, and current systems are not consistently designed to preserve identity, intent, or context in a machine-verifiable format.
The practical consequence of this gap is that liability increasingly falls on merchants. Without adapted fraud filters or shared risk protocols that account for agent behaviour, merchants face a higher incidence of chargebacks, elevated false decline rates as fraud systems misclassify legitimate agent-initiated transactions, and reduced visibility into customer relationships as agents intermediate the purchase experience.
The scale of industry concern is significant. According to a study by Accenture cited in the announcement, 87% of financial institution chief technology officers and payments heads consider trust to be the primary barrier to agentic payments adoption, while 78% expect fraud to increase as the model scales.
Framework structure and capabilities
The Unified Trust Layer framework is organised around three functional capabilities. The first is agent identity validation, which focuses on normalising identity verification and behaviour patterns across the range of fragmented agentic protocols currently in use. The second is intent capture and preservation, which involves binding explicit human consent and transaction context into an immutable, replayable record at the point of initiation. The third is contextual trust signalling, through which behavioural signals from agents are aggregated across providers, networks, and protocols to support authorisation decisions and strengthen dispute defensibility.
IXOPAY contributes merchant-owned, provider-agnostic tokenisation and payment orchestration capabilities to the initiative, while Zip brings real-time behavioural and contextual intelligence developed through its consumer-facing instalment and credit products. The two companies describe the combined approach as a shift from payment orchestration to trust orchestration.
Open participation
The framework is positioned as a collaborative, open initiative rather than a proprietary standard. IXOPAY and Zip have indicated that merchants, payment networks, platforms, and other ecosystem participants are invited to contribute to its development. No timeline for publication of formal specifications or adoption milestones has been disclosed at this stage.
The initiative reflects a broader industry conversation around the readiness of payment rails for agentic use cases, a question that has gained urgency as large language model deployments move from content generation into transactional and operational contexts. Addressing liability, auditability, and interoperability at the infrastructure level represents an early attempt to establish shared standards before agentic commerce scales further.
