What is a vCon?

A vCon is a virtualized conversation.

A structured, portable, machine-readable memory object that represents a communication event.

If a conversation is an experience, a vCon is its memory.

What a vCon contains

Everything a conversation needs to be remembered correctly

Participants
Communication channels used
Interaction timing
Surrounding context
Derived insights
Governance information

An open standard, not a vendor's idea

Not just us saying this

vCon is being standardized in the open, by the same body that gave us TCP/IP and TLS, and it already has a technical spec and a foundation behind it.

IETF vCon Working Group. An active IETF working group, the same standards body behind TCP/IP and TLS, standardizing how conversational data is packaged and exchanged. datatracker.ietf.org/group/vcon →
Conserver.io. “An open IETF standard for packaging the parties, the dialog, the recording or transcript, the consent, and the analysis into one signed, portable JSON object.” conserver.io →
The vCon Foundation. “vCons must be safe to create, store and share, ensuring user consent.” Open, cryptographically verifiable, and designed to be tamper evident. thevconfoundation.com →

Why vCons matter

As AI systems participate in conversations, memory becomes essential.

A vCon is robot food for AI.

AI systems work from governed memory, not raw transcripts. That distinction is what makes a conversation usable, safely, after it ends.

Not a recording

vCons are not recordings or transcripts

A vCon is a memory container that binds structure, context, and permissions together. A call recording is raw audio. A transcript is raw text. A vCon is what a system can actually act on, safely.

Governed memory by design

Access, reuse, and retention are part of the object itself

Governance defines access permissions, reuse parameters, purpose limitations, and retention duration. It travels with the conversation, not around it.

vCons and AI communication

Systems built on vCons remember, and they are accountable for it

They remember interactions, maintain context, respect consent, and demonstrate accountability. That is the difference between an AI system that responds and one that can be trusted with a relationship.