DeepJudge, the Zurich-based enterprise search company focused on legal AI, has launched the Agent Handoff Protocol, an open protocol designed to let users move between AI products while retaining the full context of their work1,2. Harvey and Thomson Reuters are the first adopters: Harvey is building an integration that will enter beta this month, and Thomson Reuters says it will support the protocol in CoCounsel for Legal, with implementation details and timing still to come.
The protocol addresses a friction point in multi-tool legal AI workflows. When a lawyer moves from one AI system to another — for example, from Harvey to DeepJudge — the objective, supporting materials, conversation history, and prior analysis typically do not travel with them. Agent Handoff Protocol is designed to preserve that context across transitions, and in some cases enable round-trip handoffs back to the originating product.
DeepJudge drew a distinction between Agent Handoff Protocol and existing interoperability standards. The company noted that protocols such as Model Context Protocol, developed by Anthropic, allow systems to exchange information and perform discrete actions but do not carry the broader work context — the user's objective, accumulated analysis, and conversation thread — from one product to another.
"People should be able to use the AI tool that's best for what they need to do and take their context with them. That is the experience we built Agent Handoff Protocol to enable," said Yannic Kilcher, CTO and co-founder of DeepJudge.
Siva Gurumurthy, chief technology officer at Harvey, said: "The most capable legal workflows will draw on more than one specialized system. Agent Handoff Protocol lets a user move from Harvey into a partner's product and back without losing the thread of their work, while leveraging both products to their full potential".
Joel Hron, chief technology officer at Thomson Reuters, said: "Professionals will increasingly work across specialized AI experiences, and those systems need to work together without sacrificing the context, trust, and expertise required for high-stakes work. Open standards like Agent Handoff Protocol can help enable that interoperability, which is why Thomson Reuters is excited to support this effort and help shape what comes next".
Vector Wire has previously covered the growing infrastructure layer around Model Context Protocol. On August 14, Cloudflare announced new capabilities in Cloudflare One to detect, inspect, and control Model Context Protocol traffic across managed networks ctx. ANALYSIS DeepJudge's framing of Agent Handoff Protocol as complementary to — rather than competing with — Model Context Protocol positions it as a higher-level coordination layer: where Model Context Protocol handles tool-level actions and data exchange, Agent Handoff Protocol targets the session-level state that defines a user's ongoing work. The early commitment from Harvey and Thomson Reuters, two of the most prominent AI-enabled legal platforms, gives the protocol immediate distribution in a market where multi-vendor AI stacks are becoming standard.