DeepJudge releases the Agent Handoff Protocol, which Harvey will roll out in beta this month and Thomson Reuters says it will support.
DeepJudge, a Zurich-based enterprise search company, introduced the Agent Handoff Protocol (AHP) on August 13, 2026. LawNext describes it as an open protocol designed to let users move from one AI product into another, and in some cases back again, while retaining the full context of their work.
Harvey is building an integration implementing the protocol, which will enter beta this month, according to LawNext. Artificial Lawyer reports that Harvey will roll out AHP in beta this month. Thomson Reuters says it will support AHP in CoCounsel for Legal, with details on implementation and timing to come. Siva Gurumurthy, chief technology officer at Harvey, said the protocol lets a user move from Harvey into a partner's product and back without losing the thread of their work.
DeepJudge says existing protocols such as the Model Context Protocol let systems exchange information and perform specific actions, but do not carry the broader context of the work: the objective, supporting materials, conversation history and prior analysis. DeepJudge gives the example of a lawyer researching a precedent question in DeepJudge who moves into Harvey with the context in place, without re-uploading materials, recreating prompts or rebuilding prior analysis.
Yannic Kilcher, DeepJudge's CTO and co-founder, told LawNext the handoff is a button the user clicks, and that the transfer is authenticated and scoped to the user. Joel Hron, chief technology officer at Thomson Reuters, said open standards like AHP can help enable interoperability. Technical documentation is at agenthandoffprotocol.org.
