Humanos lets agents prove authorization before moving money with signed mandates

dscape · x · 2026-09-08

When an AI agent moves money or makes an irreversible record change, access control only proves it can reach the system — not that the action itself was authorized. Humanos closes that gap: a human signs a structured mandate up front covering what's allowed, with whom, up to what limit, and for how long. Before the agent acts, the executing system calls humanos.verify(); every outcome is deterministic and leaves a portable receipt proving later exactly why the action was allowed. The team will be at Champalimaud, Lisbon, September 23–24.

Original post →

More from coding & agent

coding & agent channel →