The federal ban on a Federal Reserve central bank digital currency (CBDC) takes effect today, July 10, 2026. With public alternatives removed by H.R. 6644, private stablecoins have become the default digital dollar for agentic commerce. This shift occurs just eight days before the July 18, 2026, deadline for the GENIUS Act, a statute intended to provide regulatory clarity that has instead left a significant void.
The GENIUS Act, enacted on July 18, 2025, provided regulators one year to establish a framework for stablecoin issuers. While the OCC, FDIC, and Treasury have released Notices of Proposed Rulemaking (NPRM) this spring, the Federal Reserve has not published a standalone NPRM. The Fed’s silence is notable, particularly regarding the missing backup enforcement authority under Section 7(e)(1)(B) and the absence of anti-tying provisions under Section 4(a)(8)(B).
The legislation focuses exclusively on issuers, leaving the mechanics of AI agent payments entirely unaddressed. There is no framework for authorization, liability, or consent for autonomous systems. This creates a structural mismatch: current payment infrastructure relies on deterministic requirements, while AI agents operate on probabilistic logic.
Regulation E remains ambiguous on whether an AI agent’s authorization constitutes valid consumer consent. This leaves firms deploying agentic commerce solutions in a state of legal vulnerability. In the absence of federal guidance, the legal system is relying on the principal-agent doctrine, which places liability for an agent’s autonomous acts on the human or entity that deployed it.
International regulators are observing these limitations. In a speech at the ECB Sintra conference in June 2026, Bank of England official Sarah Breeden noted:
Our frameworks were not built to contemplate autonomous agents, and relying on a human in the loop for all agent actions is unlikely to be realistic.
The International Monetary Fund has echoed this concern. In a recent note, Davidovic and Tourpe (IMF NOTE/2026/004) emphasized the core tension:
Regulators will increasingly need to ensure that the probabilistic behaviour of agentic artificial intelligence can be reconciled with the deterministic requirements of payments infrastructure.
Meanwhile, an accelerating three-rail war is defining the market. Three distinct architectures are competing for dominance: x402, which utilizes crypto-native stablecoin settlement on Base; Google UCP, which leverages card-rails via Shopify; and Mastercard AP4M, which supports multi-rail payments including stablecoins. Regardless of the rail, merchants remain fully liable for chargebacks from AI-initiated transactions, with Visa VAMP and Mastercard ECM monitoring protocols applying as they would to human-initiated payments.
Consumer behavior remains cautious. Data from Product.ai indicates that only 14% of consumers trust AI to execute purchases without verification, while 86% require human verification before a transaction is finalized. This trust ceiling suggests that the legal and social frameworks required for autonomous commerce are still in their infancy.
As the July 18 deadline passes, market participants should monitor several developments. First, whether the Federal Reserve addresses the missing enforcement and anti-tying provisions. Second, the upcoming HM Treasury consultation on adapting payment services regulation for AI agents, which may serve as a blueprint for future policy. Finally, observers should watch how the three competing architectures evolve their internal liability and consent protocols in the absence of federal mandates. The gap between consumer behavior and existing legal protections is widening, and the regulatory response will define the next phase of agentic commerce.