Skip to main content
elan1 Govern

AI governance

Decides before the agent acts.

Your rules run before any agent's tool call does. Allow, ask a named person, or refuse, with a signed, hash-chained receipt for each. Every action, answerable.

Book a demoSee a governed action
govern.elan1.ai/approvals

Approvals

3 waiting

  • Pay Sharma Traders · ₹2,00,000

    AP agent · rule PAY-02: moves_money over ₹50,000 · 21 h 14 m left

    Held for finance head

  • Change bank details · Vendor 0412

    Refused · rule 4.2
Approved by Asha R.Receipt chain verified · 4,097
Approvals screen, 3 waiting: a ₹2,00,000 payment to Sharma Traders held for the finance head under rule PAY-02, with 21 h 14 m left; a bank detail change for Vendor 0412 refused under rule 4.2. Asha R. has approved another request, and the receipt chain of 4,097 is verified. Illustrative data.

What is elan1 Govern?

elan1 Govern is AI governance that decides before an agent's tool call runs: allow, ask a named person or refuse. Your signed rulebook runs first, anything not clearly allowed is refused, and every outcome, refusals included, leaves a signed, hash-chained receipt that an auditor can verify offline with one command.

Updated

The challenge

Most AI tools watch. Few stop anything.

Tracing and dashboards tell you what an agent did. By then the payment has gone. The moment that matters is the write, and something has to stand in front of it.

An audit lead checks a document line by line with a ruler at a high-rise office window

  • 01

    Watching is not governing.

    A trace cannot prevent an incident.

  • 02

    The actor narrates itself.

    A record inside the agent's own product is self-attestation.

  • 03

    No one can stop it.

    60% of enterprises cannot stop a misbehaving agent.

    CSA and Token Security, April 2026

How it works

Ask. Check. Approve. Act. Record.

  1. 01

    Ask.

    The agent's tool call comes to Govern first.

  2. 02

    Check.

    Your signed rulebook runs. Anything not clearly allowed is refused, and an error refuses.

  3. 03

    Approve.

    Risky actions wait for a named person. If nobody answers in 24 hours, the action is refused.

  4. 04

    Act.

    Only the allowed action runs. On the governed route, the agent never holds the credentials.

  5. 05

    Record.

    A signed, hash-chained receipt for every outcome, refusals included.

Take a look inside.

Rules your finance head can read.

  • One signed rulebook per agent, versioned
  • Rules match risk tags such as moves_money
  • Test a change on past actions first
govern.elan1.ai/rulebooks/ap-agent

Rulebook · AP agent

v14 · signed

  • PAY-02 · moves_money over ₹50,000

    Ask: finance head · expires after 24 h, then refuses

    Ask a person

  • 4.2 · Change bank details

    Never by an agent

    Refuse

  • Default · anything not clearly allowed

    An error refuses too

    Refuse

Test v15 on past actions before it is signed

Rulebook for the AP agent, version 14, signed: payments that move money over ₹50,000 ask the finance head; bank detail changes are refused under rule 4.2; anything not clearly allowed is refused. Illustrative data.
Take the tour

By the numbers.

A vendor paid only for the yes has a reason to say yes.

A governance lead holds a printed rule sheet against a signed record with a crimson stamp, comparing them under a reading lamp in an archive room

Where it works.

All industries

How it fits

The one layer that allows.

Act, Propel, Resolve, Decide and Pulse propose. Run carries. elan1 Ontology defines what may change. Only Govern allows, for our agents and for yours.

Applications

elan1 Actelan1 Propelelan1 Resolveelan1 Pulse

Foundation

elan1 Ontologyelan1 Decideelan1 Governelan1 Run
elan1 Govern on the elan1 platform.

Questions

elan1 Govern, answered plainly.

Ask us anything

AI agent governance decides whether an agent's action may run before it runs, rather than watching it afterwards. In elan1 Govern, each tool call is checked against your signed rulebook and gets one of three outcomes: allow, ask a named person or refuse. Anything not clearly allowed is refused, and every outcome leaves a signed receipt.

Monitoring watches after the fact. Govern stands in front of the action. Keep the monitoring. Add the gate.

Saying no takes the same checks, and leaves the same receipt, as saying yes. Pricing it removes our reason to prefer the yes.

Yes. Export the pack and run one command offline. Nothing phones home.

Yes. With elan1 Govern, the agent's tool call comes to Govern first, and only the allowed action runs. Risky actions wait for a named person, and if nobody answers in 24 hours, the action is refused. An error refuses too. On the governed route, the agent never holds the credentials, and typing STOP pauses every agent.

You can build an agent. The hard part is around it: rules that refuse by default, approvals that expire, credentials the agent never holds, and receipts your auditor verifies without trusting you. That part is the product.

Start with one process.

Bring one agent and one rule. We'll show you the receipt.

Book a demo
  • Allowed · Approved by Asha R.
  • Held for CFO
  • Refused · rule 4.2
Illustrative data.

Start with one process.

Add the next when it holds.

Book a demo
Approvals · Kolam ComponentsIllustrative
  • Allowed · Approved by Asha R.
  • Held for CFO
  • Refused · rule 4.2
RCPT-0042 · signed · chained
A product view with three decisions: allowed and approved by Asha R., held for the CFO, and refused under rule 4.2. Illustrative data.