Underwriting infrastructure for AI agents
The data layer between AI agents and the people who insure them.
Companies are deploying agents faster than their insurers can price them. Locard gives underwriters a standard way to see what those agents actually do — and what it would cost when one gets it wrong.
The questions nobody can answer
An insurer writing a policy today has no way to answer three basic questions about the company in front of them. Nothing in the submission process asks. Nothing in the data would answer.
- What agents does this company actually run?
- What systems and permissions can those agents reach?
- How often do they act without a human in the loop?
Three things.
Connect
One integration, any agent stack. Companies send agent data to their insurer through Locard instead of building a bespoke export for every carrier that asks.
Translate
Raw agent telemetry is not an underwriting input. We turn it into the variables risk and actuarial teams already model against.
Distinguish
Two companies running agents are not the same risk. A scoped deployment with a human on the escalation path is not the one that gave an agent a corporate card. Locard makes the difference legible enough to price.
Who we work with
You're being asked to price something you can't see.
Carriers, MGAs and brokers underwriting companies that have put agents into production — and who would rather model that exposure than exclude it.
You'd rather be measured than guessed at.
If your agents are well governed, that's currently invisible to the person setting your premium. Locard is how it stops being invisible.
What's with the name?
Edmond Locard was a French forensic scientist. His exchange principle held that every contact leaves a trace. Agents are no exception — the traces are already there. Nobody has been collecting them in a form an underwriter can use.
"Every contact leaves a trace."
Locard's exchange principle — forensic science, c. 1910
Insurance can save the world. It just needs the data.
We're onboarding early partners on both sides of the market. Tell us which one you're on.