Solutions — start from the outcome you need
You did not come here for a runtime diagram. You came because an AI agent is about to do something consequential — move money, close an alert, sign a document — and someone will ask who let it. KYE Protocol™ starts from that outcome. Tell us the problem you have to solve today, and each path below routes you to the exact KYE Protocol™ capability that solves it — then proves it to your auditor. Most teams cut authority evidence-gathering from days to minutes.
Four outcomes, one authority engine
Every solution below is a different question a buyer asks, answered by the same underlying engine: KYE Protocol™ checks whether an action is admissible, records the decision, and lets anyone replay it. Pick the outcome that matches your week — you do not need to understand the whole protocol to adopt one.
AI agent auditability
Your regulator or board wants a record of every consequential action an agent took — and who authorised it. KYE Protocol™ attaches a verifiable decision to each action, so an AML analyst can show, in minutes, why an alert was closed.
AI agent auditability →Runtime approval
Some actions must pause for a human. KYE Protocol™ enforces Action Admissibility™ at the moment of the call — allow, deny, or require sign-off — with two-person approval on irreversibles, mapped to EU AI Act™ Article 14 human oversight.
Runtime approval with GovernedUI™ →Evidence replay
An assertion is not evidence. KYE Protocol™ seals each decision into an Evidence Pack™ that a third party can replay from public keys alone — the Replay-Proof™ record ISO/IEC 42001 and SOC 2 auditors accept without trusting your word.
Evidence Pack™ →Consequential-action governance
Above a threshold, an agent action needs the full lifecycle — purpose, scope, decision, evidence, finality. KYE Protocol™ governs that end to end, mapped to NIST AI RMF and DORA, so capability never silently becomes authority.
The authority lifecycle →How a solution becomes an adoption
Each path starts as a scoped pilot and graduates to production on the same engine — no rebuild. A CISO evaluating auditability, a General Counsel checking human-oversight, and a builder wiring approvals all land on one governed runtime, so your first outcome does not become a dead end.
- Evaluate — run KYE Protocol™ in Shadow Mode™ beside your agents. It records what it would have decided, with zero blast radius, across all 163 mapped frameworks.
- Approve — turn on runtime approval for the actions that matter, keeping a human in control of AI authority.
- Prove — hand your auditor a replayable Evidence Pack™ instead of a spreadsheet, cutting exam-prep from days to minutes.
Not sure which outcome is yours? Start from your role or talk to us about a pilot.