Most agent protocol work so far has answered two questions: who is this agent, and how does it pay. The American Arbitration Association just launched something aimed at a third question nobody had standardized: under what legal terms did this transaction happen, and what happens when it goes wrong. The AAA, working with Integra Ledger as lead technical steward, introduced the Legal Context Protocol (LCP) this summer alongside a coalition that includes Google, IBM, Circle, Wayfair, UiPath, and several blockchain foundations. The pitch is explicit about the gap it's filling in the emerging agent stack: payment protocols answer what was paid; identity and coordination frameworks answer who acted; LCP answers under what terms, governed by what law, and with what recourse. Concretely, the protocol is described as making legal terms, consent, and dispute resolution discoverable and verifiable when AI agents transact on behalf of people and organizations. That's a narrower and more legally specific claim than most identity or attestation frameworks make. Where a delegation credential or a signed receipt proves an agent had authority to act, LCP is trying to attach the actual contractual terms — governing law, dispute process, recourse mechanism — to the transaction itself, in a form an agent (or a court) can check. This matters because the accountability gap in agent commerce isn't just "was this agent authorized." It's "if this agent's purchase, refund, or contract execution goes wrong, whose arbitration clause applies, and can anyone prove what was agreed to at the moment the agent acted." Payment rails and identity credentials don't answer that. A protocol backed by the country's largest commercial arbitration body is a signal that legal recourse infrastructure — not just cryptographic trust — is becoming a recognized missing layer in agent-to-agent transactions. The coalition list is notable for its breadth: enterprise cloud players (Google, IBM), payment/stablecoin infrastructure (Circle), a major retailer (Wayfair), automation vendors (UiPath), and a cluster of blockchain foundations (Stellar, Cardano, Hedera, Aptos, Sei). That spread suggests LCP is being pitched less as a niche legal-tech spec and more as a layer meant to sit underneath any agent commerce stack, regardless of which blockchain or payment rail carries the transaction. What's untested is adoption mechanics: whether agents will actually query and honor LCP terms before transacting, and whether arbitration outcomes under the protocol hold up when tested in a real dispute. A standard for discoverable terms is not the same as enforcement. But naming the gap — recourse, not just identity or payment — is itself the news here.