Salesforce 论文发现跨方智能体不会主动核实对方说法
New Salesforce paper finds that agents acting for different parties never verify each other's claims unprompted, so teams inherit their weakest agent's errors.
Salesforce 新论文发现，代表不同利益方的智能体在无提示时不会核实同伴的说法，团队因此会继承最弱智能体的错误。研究在独立沙箱中用 Claude Code、Codex 和 Cursor 跑了 488 个场景、超过 2500 次智能体间会话，从轨迹中归纳出 20 种失败模式。没有智能体重新获取文档来验证同伴的事实性说法，且智能体主要在提示要求时才联系同伴；在一例服务纠纷中，24000 美元的和解方案通过了全部 10 项结果检查，但供应商方智能体泄露了机密信息。
正文
New Salesforce paper finds that agents acting for different parties never verify each other's claims unprompted, so teams inherit their weakest agent's errors.
Agents increasingly coordinate with agents that represent other people or companies, each with private files, tools, and goals. Existing failure studies assume 1 orchestrator or a shared workspace.
They ran over 2,500 agent-to-agent sessions across 488 scenarios, using Claude Code, Codex, and Cursor in separate sandboxes. They catalogued 20 failure modes from the traces.
No agent re-fetched a document to test a peer's factual claim, and agents contacted peers mainly when the prompt told them to. In 1 service-level dispute, a $24,000 settlement passed all 10 outcome checks while the vendor agent leaked confidential details.
If your agent talks to outside agents, build checks into the protocol and score each agent against the party it serves, not the team result.